---
title: Messages
url: https://platform.claude.com/docs/en/api/php/messages
---

# Messages

## Create a Message

`$client->messages->create(int maxTokens, list<MessageParam> messages, Model model, ?CacheControlEphemeral cacheControl, ?MessageCreateParamsContainer container, ?DiagnosticsParam diagnostics, ?string inferenceGeo, ?Metadata metadata, ?OutputConfig outputConfig, ?ServiceTier serviceTier, ?Speed speed, ?list<string> stopSequences, ?System system, ?ThinkingConfigParam thinking, ?ToolChoice toolChoice, ?list<ToolUnion> tools, ?string userProfileID, ?string workspaceID): Message`

**POST** `/v1/messages`

Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.

Learn more about the Messages API in our [user guide](https://platform.claude.com/docs/en/get-started?sdk_languages=php)

### Parameters

- `maxTokens: int`

  The maximum number of tokens to generate before stopping.

  Note that our models may stop _before_ reaching this maximum. This parameter only specifies the absolute maximum number of tokens to generate.

  Set to `0` to populate the [prompt cache](https://platform.claude.com/docs/en/build-with-claude/prompt-caching?sdk_languages=php#pre-warming-the-cache) without generating a response.

  Different models have different maximum values for this parameter.  See [models](https://platform.claude.com/docs/en/about-claude/models/overview?sdk_languages=php) for details.

- `messages: list<MessageParam>`

  Input messages.

  Our models are trained to operate on alternating `user` and `assistant` conversational turns. When creating a new `Message`, you specify the prior conversational turns with the `messages` parameter, and the model then generates the next `Message` in the conversation. Consecutive `user` or `assistant` turns in your request will be combined into a single turn.

  Each input message must be an object with a `role` and `content`. You can specify a single `user`-role message, or you can include multiple `user` and `assistant` messages.

  If the final message uses the `assistant` role, the response content will continue immediately from the content in that message. This can be used to constrain part of the model's response. This is called prefill. On models that don't support prefill, creating a message that ends with a partial `assistant` response returns a 400 error. See [Prefill not supported](https://platform.claude.com/docs/en/api/errors?sdk_languages=php#prefill-not-supported).

  Example with a single `user` message:

  ```json
  [{"role": "user", "content": "Hello, Claude"}]
  ```

  Example with multiple conversational turns:

  ```json
  [
    {"role": "user", "content": "Hello there."},
    {"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},
    {"role": "user", "content": "Can you explain LLMs in plain English?"},
  ]
  ```

  Example with a partially-filled response from Claude, for models that support prefill:

  ```json
  [
    {"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
    {"role": "assistant", "content": "The best answer is ("},
  ]
  ```

  Each input message `content` may be either a single `string` or an array of content blocks, where each block has a specific `type`. Using a `string` for `content` is shorthand for an array of one content block of type `"text"`. The following input messages are equivalent:

  ```json
  {"role": "user", "content": "Hello, Claude"}
  ```

  ```json
  {"role": "user", "content": [{"type": "text", "text": "Hello, Claude"}]}
  ```

  See [input examples](https://platform.claude.com/docs/en/build-with-claude/working-with-messages?sdk_languages=php).

  Note that if you want to include a [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices?sdk_languages=php#give-claude-a-role), you can use the top-level `system` parameter.

  There is a limit of 100,000 messages in a single request.

- `model: Model`

  The model that will complete your prompt.

  See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.

- `cacheControl?: optional CacheControlEphemeral`

  Top-level cache control automatically applies a cache_control marker to the last cacheable block in the request.

- `container?: optional MessageCreateParamsContainer`

  Container identifier for reuse across requests.

- `diagnostics?: optional DiagnosticsParam`

  Request-level diagnostics. Supply `previous_message_id` to have the response include `diagnostics.cache_miss_reason` explaining any prompt-cache divergence from that prior request.

- `inferenceGeo?: optional string`

  Specifies the geographic region for inference processing. If not specified, the workspace's `default_inference_geo` is used.

- `metadata?: optional Metadata`

  An object describing metadata about the request.

- `outputConfig?: optional OutputConfig`

  Configuration options for the model's output, such as the output format.

- `serviceTier?: optional ServiceTier`

  Determines whether to use priority capacity (if available) or standard capacity for this request.

  Anthropic offers different levels of service for your API requests. See [service-tiers](https://platform.claude.com/docs/en/api/service-tiers?sdk_languages=php) for details.

- `speed?: optional Speed`

  The inference speed mode for this request. `"fast"` enables high output-tokens-per-second inference.

  Some models require the `fast-mode-2026-02-01` beta header for any `speed` value. See [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode?sdk_languages=php) for which models.

- `stopSequences?: optional list<string>`

  Custom text sequences that will cause the model to stop generating.

  Our models will normally stop when they have naturally completed their turn, which will result in a response `stop_reason` of `"end_turn"`.

  If you want the model to stop generating when it encounters custom strings of text, you can use the `stop_sequences` parameter. If the model encounters one of the custom sequences, the response `stop_reason` value will be `"stop_sequence"` and the response `stop_sequence` value will contain the matched stop sequence.

- `system?: optional System`

  System prompt.

  A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role. See our [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices?sdk_languages=php#give-claude-a-role).

- `thinking?: optional ThinkingConfigParam`

  Configuration for Claude's thinking.

  With `{"type": "adaptive"}`, Claude decides when and how much to think. With `{"type": "enabled"}` (manual extended thinking), you set a `budget_tokens` of at least 1,024. Thinking tokens count toward your `max_tokens` limit.

  Which `type` values are accepted, and what happens when you omit `thinking`, depend on the model. See [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#configuring-thinking) for each model's behavior.

- `toolChoice?: optional ToolChoice`

  How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.

- `tools?: optional list<ToolUnion>`

  Definitions of tools that the model may use.

  If you include `tools` in your API request, the model may return `tool_use` content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using `tool_result` content blocks.

  There are two types of tools: **client tools** and **server tools**. The behavior described below applies to client tools. For [server tools](https://platform.claude.com/docs/en/agents-and-tools/tool-use/server-tools?sdk_languages=php), see their individual documentation as each has its own behavior (e.g., the [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool?sdk_languages=php)).

  Each tool definition includes:

  * `name`: Name of the tool.
  * `description`: Optional, but strongly-recommended description of the tool.
  * `input_schema`: [JSON schema](https://json-schema.org/draft/2020-12) for the tool `input` shape that the model will produce in `tool_use` output content blocks.

  For example, if you defined `tools` as:

  ```json
  [
    {
      "name": "get_stock_price",
      "description": "Get the current stock price for a given ticker symbol.",
      "input_schema": {
        "type": "object",
        "properties": {
          "ticker": {
            "type": "string",
            "description": "The stock ticker symbol, e.g. AAPL for Apple Inc."
          }
        },
        "required": ["ticker"]
      }
    }
  ]
  ```

  And then asked the model "What's the S&P 500 at today?", the model might produce `tool_use` content blocks in the response like this:

  ```json
  [
    {
      "type": "tool_use",
      "id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
      "name": "get_stock_price",
      "input": { "ticker": "^GSPC" }
    }
  ]
  ```

  You might then run your `get_stock_price` tool with `{"ticker": "^GSPC"}` as an input, and return the following back to the model in a subsequent `user` message:

  ```json
  [
    {
      "type": "tool_result",
      "tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
      "content": "259.75 USD"
    }
  ]
  ```

  Tools can be used for workflows that include running client-side tools and functions, or more generally whenever you want the model to produce a particular JSON structure of output.

  See our [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview?sdk_languages=php) for more details.

- `userProfileID?: optional string` (header parameter)

  The user profile ID to attribute this request to. Use when acting on behalf of a party other than your organization.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

### Returns

- `class Message`

  - `"message" type`

    Object type.

    For Messages, this is always `"message"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?Container container`

    Information about the container used in this request.

    This will be non-null if a container tool (e.g. code execution) was used.

  - `list<ContentBlock> content`

    Content generated by the model.

    This is an array of content blocks, each of which has a `type` that determines its shape.

    Example:

    ```json
    [{"type": "text", "text": "Hi, I'm Claude."}]
    ```

    If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.

    For example, if the input `messages` were:

    ```json
    [
      {"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
      {"role": "assistant", "content": "The best answer is ("}
    ]
    ```

    Then the response `content` might be:

    ```json
    [{"type": "text", "text": "B)"}]
    ```

  - `?Diagnostics diagnostics`

    Request-level diagnostics. `null` when the request did not supply `diagnostics`, or when it did and no prompt-cache divergence was detected.

  - `Model model`

    The model that will complete your prompt.

    See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.

  - `"assistant" role`

    Conversational role of the generated message.

    This will always be `"assistant"`.

  - `?RefusalStopDetails stopDetails`

    Structured information about why model output stopped.

    This is `null` when the `stop_reason` has no additional detail to report.

  - `?StopReason stopReason`

    The reason that we stopped.

    This may be one the following values:

    * `"end_turn"`: the model reached a natural stopping point
    * `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum
    * `"stop_sequence"`: one of your provided custom `stop_sequences` was generated
    * `"tool_use"`: the model invoked one or more tools
    * `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.
    * `"refusal"`: when streaming classifiers intervene to handle potential policy violations
    * `"model_context_window_exceeded"`: we exceeded the model's context window

    In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.

  - `?string stopSequence`

    Which custom stop sequence was generated, if any.

    This value will be a non-null string if one of your custom stop sequences was generated.

  - `Usage usage`

    Billing and rate-limit usage.

    Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.

    Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.

    For example, `output_tokens` will be non-zero, even for an empty string response from Claude.

    Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.

- `class RawMessageStreamEvent`

  - `class RawMessageStartEvent`

    - `"message_start" type`

    - `Message message`

  - `class RawMessageDeltaEvent`

    - `"message_delta" type`

    - `Delta delta`

    - `MessageDeltaUsage usage`

      Billing and rate-limit usage.

      Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.

      Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.

      For example, `output_tokens` will be non-zero, even for an empty string response from Claude.

      Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.

  - `class RawMessageStopEvent`

    - `"message_stop" type`

  - `class RawContentBlockStartEvent`

    - `"content_block_start" type`

    - `ContentBlock contentBlock`

    - `int index`

  - `class RawContentBlockDeltaEvent`

    - `"content_block_delta" type`

    - `RawContentBlockDelta delta`

    - `int index`

  - `class RawContentBlockStopEvent`

    - `"content_block_stop" type`

    - `int index`

### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;
use Anthropic\Messages\Model;

$client = new Client(apiKey: 'my-anthropic-api-key');

$message = $client->messages->create(
  maxTokens: 1024,
  messages: [['content' => 'Hello, world', 'role' => 'user']],
  model: Model::CLAUDE_OPUS_5_5,
  cacheControl: ['type' => 'ephemeral', 'ttl' => '5m'],
  container: [
    'id' => 'id',
    'skills' => [
      ['skillID' => 'pdf', 'type' => 'anthropic', 'version' => 'latest']
    ],
  ],
  diagnostics: ['previousMessageID' => 'previous_message_id'],
  inferenceGeo: 'inference_geo',
  metadata: ['userID' => '13803d75-b4b5-4c3e-b2a2-6f21399b021b'],
  outputConfig: [
    'effort' => 'low',
    'format' => ['schema' => ['foo' => 'bar'], 'type' => 'json_schema'],
  ],
  serviceTier: 'auto',
  speed: 'standard',
  stopSequences: ['string'],
  system: [
    [
      'text' => 'Today\'s date is 2024-06-01.',
      'type' => 'text',
      'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
      'citations' => [
        [
          'citedText' => 'The grass is green. The sky is blue.',
          'documentIndex' => 0,
          'documentTitle' => 'x',
          'endCharIndex' => 0,
          'startCharIndex' => 0,
          'type' => 'char_location',
        ],
      ],
    ],
  ],
  thinking: ['type' => 'adaptive', 'display' => 'summarized'],
  toolChoice: ['type' => 'auto', 'disableParallelToolUse' => true],
  tools: [
    [
      'inputSchema' => [
        'type' => 'object',
        'properties' => ['location' => 'bar', 'unit' => 'bar'],
        'required' => ['location'],
      ],
      'name' => 'name',
      'allowedCallers' => ['direct'],
      'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
      'deferLoading' => true,
      'description' => 'Get the current weather in a given location',
      'eagerInputStreaming' => true,
      'inputExamples' => [['foo' => 'bar']],
      'strict' => true,
      'type' => 'custom',
    ],
  ],
  userProfileID: 'anthropic-user-profile-id',
  workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy',
);

var_dump($message);
```

#### Response (200)

```json
{
  "id": "msg_013Zva2CMHLNnXjNJJKqJ2EF",
  "container": {
    "id": "container_011CpZohnwH4vuy7gazohgSP",
    "expires_at": "2019-12-27T18:11:19.117Z",
    "skills": [
      {
        "skill_id": "pdf",
        "type": "anthropic",
        "version": "latest"
      }
    ]
  },
  "content": [
    {
      "citations": [
        {
          "cited_text": "The grass is green. The sky is blue.",
          "document_index": 0,
          "document_title": "My Document",
          "end_char_index": 0,
          "file_id": "file_011CNha8iCJcU1wXNR6q4V8w",
          "start_char_index": 0,
          "type": "char_location"
        }
      ],
      "text": "Hi! My name is Claude.",
      "type": "text"
    }
  ],
  "diagnostics": {
    "cache_miss_reason": {
      "cache_missed_input_tokens": 0,
      "type": "model_changed"
    }
  },
  "model": "claude-opus-5-5",
  "role": "assistant",
  "stop_details": {
    "category": "cyber",
    "explanation": "This request was declined because it conflicts with Anthropic's Usage Policy.",
    "type": "refusal"
  },
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "type": "message",
  "usage": {
    "cache_creation": {
      "ephemeral_1h_input_tokens": 0,
      "ephemeral_5m_input_tokens": 0
    },
    "cache_creation_input_tokens": 2051,
    "cache_read_input_tokens": 2051,
    "inference_geo": "global",
    "input_tokens": 2095,
    "output_tokens": 503,
    "output_tokens_details": {
      "thinking_tokens": 0
    },
    "server_tool_use": {
      "web_fetch_requests": 2,
      "web_search_requests": 0
    },
    "service_tier": "standard",
    "speed": "standard"
  }
}
```

## Count tokens in a Message

`$client->messages->countTokens(list<MessageParam> messages, Model model, ?CacheControlEphemeral cacheControl, ?OutputConfig outputConfig, ?Speed speed, ?System system, ?ThinkingConfigParam thinking, ?ToolChoice toolChoice, ?list<MessageCountTokensTool> tools, ?string userProfileID, ?string workspaceID): MessageTokensCount`

**POST** `/v1/messages/count_tokens`

Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.

Learn more about token counting in our [user guide](https://platform.claude.com/docs/en/build-with-claude/token-counting?sdk_languages=php)

### Parameters

- `messages: list<MessageParam>`

  Input messages.

  Our models are trained to operate on alternating `user` and `assistant` conversational turns. When creating a new `Message`, you specify the prior conversational turns with the `messages` parameter, and the model then generates the next `Message` in the conversation. Consecutive `user` or `assistant` turns in your request will be combined into a single turn.

  Each input message must be an object with a `role` and `content`. You can specify a single `user`-role message, or you can include multiple `user` and `assistant` messages.

  If the final message uses the `assistant` role, the response content will continue immediately from the content in that message. This can be used to constrain part of the model's response. This is called prefill. On models that don't support prefill, creating a message that ends with a partial `assistant` response returns a 400 error. See [Prefill not supported](https://platform.claude.com/docs/en/api/errors?sdk_languages=php#prefill-not-supported).

  Example with a single `user` message:

  ```json
  [{"role": "user", "content": "Hello, Claude"}]
  ```

  Example with multiple conversational turns:

  ```json
  [
    {"role": "user", "content": "Hello there."},
    {"role": "assistant", "content": "Hi, I'm Claude. How can I help you?"},
    {"role": "user", "content": "Can you explain LLMs in plain English?"},
  ]
  ```

  Example with a partially-filled response from Claude, for models that support prefill:

  ```json
  [
    {"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
    {"role": "assistant", "content": "The best answer is ("},
  ]
  ```

  Each input message `content` may be either a single `string` or an array of content blocks, where each block has a specific `type`. Using a `string` for `content` is shorthand for an array of one content block of type `"text"`. The following input messages are equivalent:

  ```json
  {"role": "user", "content": "Hello, Claude"}
  ```

  ```json
  {"role": "user", "content": [{"type": "text", "text": "Hello, Claude"}]}
  ```

  See [input examples](https://platform.claude.com/docs/en/build-with-claude/working-with-messages?sdk_languages=php).

  Note that if you want to include a [system prompt](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices?sdk_languages=php#give-claude-a-role), you can use the top-level `system` parameter.

  There is a limit of 100,000 messages in a single request.

- `model: Model`

  The model that will complete your prompt.

  See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.

- `cacheControl?: optional CacheControlEphemeral`

  Top-level cache control automatically applies a cache_control marker to the last cacheable block in the request.

- `outputConfig?: optional OutputConfig`

  Configuration options for the model's output, such as the output format.

- `speed?: optional Speed`

  The inference speed mode for this request. `"fast"` enables high output-tokens-per-second inference.

  Some models require the `fast-mode-2026-02-01` beta header for any `speed` value. See [fast mode](https://platform.claude.com/docs/en/build-with-claude/fast-mode?sdk_languages=php) for which models.

- `system?: optional System`

  System prompt.

  A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role. See our [guide to system prompts](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/claude-prompting-best-practices?sdk_languages=php#give-claude-a-role).

- `thinking?: optional ThinkingConfigParam`

  Configuration for Claude's thinking.

  With `{"type": "adaptive"}`, Claude decides when and how much to think. With `{"type": "enabled"}` (manual extended thinking), you set a `budget_tokens` of at least 1,024. Thinking tokens count toward your `max_tokens` limit.

  Which `type` values are accepted, and what happens when you omit `thinking`, depend on the model. See [thinking](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#configuring-thinking) for each model's behavior.

- `toolChoice?: optional ToolChoice`

  How the model should use the provided tools. The model can use a specific tool, any available tool, decide by itself, or not use tools at all.

- `tools?: optional list<MessageCountTokensTool>`

  Definitions of tools that the model may use.

  If you include `tools` in your API request, the model may return `tool_use` content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using `tool_result` content blocks.

  There are two types of tools: **client tools** and **server tools**. The behavior described below applies to client tools. For [server tools](https://platform.claude.com/docs/en/agents-and-tools/tool-use/server-tools?sdk_languages=php), see their individual documentation as each has its own behavior (e.g., the [web search tool](https://platform.claude.com/docs/en/agents-and-tools/tool-use/web-search-tool?sdk_languages=php)).

  Each tool definition includes:

  * `name`: Name of the tool.
  * `description`: Optional, but strongly-recommended description of the tool.
  * `input_schema`: [JSON schema](https://json-schema.org/draft/2020-12) for the tool `input` shape that the model will produce in `tool_use` output content blocks.

  For example, if you defined `tools` as:

  ```json
  [
    {
      "name": "get_stock_price",
      "description": "Get the current stock price for a given ticker symbol.",
      "input_schema": {
        "type": "object",
        "properties": {
          "ticker": {
            "type": "string",
            "description": "The stock ticker symbol, e.g. AAPL for Apple Inc."
          }
        },
        "required": ["ticker"]
      }
    }
  ]
  ```

  And then asked the model "What's the S&P 500 at today?", the model might produce `tool_use` content blocks in the response like this:

  ```json
  [
    {
      "type": "tool_use",
      "id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
      "name": "get_stock_price",
      "input": { "ticker": "^GSPC" }
    }
  ]
  ```

  You might then run your `get_stock_price` tool with `{"ticker": "^GSPC"}` as an input, and return the following back to the model in a subsequent `user` message:

  ```json
  [
    {
      "type": "tool_result",
      "tool_use_id": "toolu_01D7FLrfh4GYq7yT1ULFeyMV",
      "content": "259.75 USD"
    }
  ]
  ```

  Tools can be used for workflows that include running client-side tools and functions, or more generally whenever you want the model to produce a particular JSON structure of output.

  See our [guide](https://platform.claude.com/docs/en/agents-and-tools/tool-use/overview?sdk_languages=php) for more details.

- `userProfileID?: optional string` (header parameter)

  The user profile ID to attribute this request to. Use when acting on behalf of a party other than your organization.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

### Returns

- `class MessageTokensCount`

  - `int inputTokens`

    The total number of tokens across the provided list of messages, system prompt, and tools.

### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;
use Anthropic\Messages\Model;

$client = new Client(apiKey: 'my-anthropic-api-key');

$messageTokensCount = $client->messages->countTokens(
  messages: [['content' => 'Hello, world', 'role' => 'user']],
  model: Model::CLAUDE_OPUS_5_5,
  cacheControl: ['type' => 'ephemeral', 'ttl' => '5m'],
  outputConfig: [
    'effort' => 'low',
    'format' => ['schema' => ['foo' => 'bar'], 'type' => 'json_schema'],
  ],
  speed: 'standard',
  system: [
    [
      'text' => 'Today\'s date is 2024-06-01.',
      'type' => 'text',
      'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
      'citations' => [
        [
          'citedText' => 'The grass is green. The sky is blue.',
          'documentIndex' => 0,
          'documentTitle' => 'x',
          'endCharIndex' => 0,
          'startCharIndex' => 0,
          'type' => 'char_location',
        ],
      ],
    ],
  ],
  thinking: ['type' => 'adaptive', 'display' => 'summarized'],
  toolChoice: ['type' => 'auto', 'disableParallelToolUse' => true],
  tools: [
    [
      'inputSchema' => [
        'type' => 'object',
        'properties' => ['location' => 'bar', 'unit' => 'bar'],
        'required' => ['location'],
      ],
      'name' => 'name',
      'allowedCallers' => ['direct'],
      'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
      'deferLoading' => true,
      'description' => 'Get the current weather in a given location',
      'eagerInputStreaming' => true,
      'inputExamples' => [['foo' => 'bar']],
      'strict' => true,
      'type' => 'custom',
    ],
  ],
  userProfileID: 'anthropic-user-profile-id',
  workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy',
);

var_dump($messageTokensCount);
```

#### Response (200)

```json
{
  "input_tokens": 2095
}
```

## Domain types

### Assistant Message Param

- `class AssistantMessageParam`

  - `MessageContentParam content`

  - `"assistant" role`

### Base64 Image Source

- `class Base64ImageSource`

  - `"base64" type`

  - `string data`

  - `MediaType mediaType`

### Base64 PDF Source

- `class Base64PDFSource`

  - `"base64" type`

  - `string data`

  - `"application/pdf" mediaType`

### Bash Code Execution Output Block

- `class BashCodeExecutionOutputBlock`

  - `"bash_code_execution_output" type`

  - `string fileID`

### Bash Code Execution Output Block Param

- `class BashCodeExecutionOutputBlockParam`

  - `"bash_code_execution_output" type`

  - `string fileID`

### Bash Code Execution Result Block

- `class BashCodeExecutionResultBlock`

  - `"bash_code_execution_result" type`

  - `list<BashCodeExecutionOutputBlock> content`

  - `int returnCode`

  - `string stderr`

  - `string stdout`

### Bash Code Execution Result Block Param

- `class BashCodeExecutionResultBlockParam`

  - `"bash_code_execution_result" type`

  - `list<BashCodeExecutionOutputBlockParam> content`

  - `int returnCode`

  - `string stderr`

  - `string stdout`

### Bash Code Execution Tool Result Block

- `class BashCodeExecutionToolResultBlock`

  - `"bash_code_execution_tool_result" type`

  - `Content content`

  - `string toolUseID`

### Bash Code Execution Tool Result Block Param

- `class BashCodeExecutionToolResultBlockParam`

  - `"bash_code_execution_tool_result" type`

  - `Content content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Bash Code Execution Tool Result Error

- `class BashCodeExecutionToolResultError`

  - `"bash_code_execution_tool_result_error" type`

  - `BashCodeExecutionToolResultErrorCode errorCode`

### Bash Code Execution Tool Result Error Code

- `enum BashCodeExecutionToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"unavailable"`

  - `"too_many_requests"`

  - `"execution_time_exceeded"`

  - `"output_file_too_large"`

### Bash Code Execution Tool Result Error Param

- `class BashCodeExecutionToolResultErrorParam`

  - `"bash_code_execution_tool_result_error" type`

  - `BashCodeExecutionToolResultErrorCode errorCode`

### Browser Click Target

- `class BrowserClickTarget`

  - `class BrowserCoordinateTarget`

    - `"coordinate" type`

    - `int x`

      Pixels from the left edge of the viewport.

    - `int y`

      Pixels from the top edge of the viewport.

  - `class BrowserRefTarget`

    - `"ref" type`

    - `string ref`

      An element reference (e.g. "ref_7") returned by a prior `read_page` or `find` result.

### Browser Close Tab Config

- `class BrowserCloseTabConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Close Tab Input

- `class BrowserCloseTabInput`

  - `string tabID`

    The tab to close.

### Browser Close Tab Tool Use Block

- `class BrowserCloseTabToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserCloseTabInput input`

    Close the tab with the given tab_id.

  - `"close_tab" name`

  - `"browser" toolsetName`

### Browser Coordinate Target

- `class BrowserCoordinateTarget`

  - `"coordinate" type`

  - `int x`

    Pixels from the left edge of the viewport.

  - `int y`

    Pixels from the top edge of the viewport.

### Browser Double Click Config

- `class BrowserDoubleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Double Click Input

- `class BrowserDoubleClickInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string modifiers`

    Optional modifier key chord to hold for the duration of this action (e.g. "shift", "ctrl+shift", "cmd+alt").

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Double Click Tool Use Block

- `class BrowserDoubleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserDoubleClickInput input`

    Double left-click at a viewport coordinate or on an element by reference.

  - `"double_click" name`

  - `"browser" toolsetName`

### Browser File Upload Config

- `class BrowserFileUploadConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser File Upload Input

- `class BrowserFileUploadInput`

  - `BrowserRefTarget target`

    An element on the page, identified by a reference from a prior `read_page` or
    `find` result. References are scoped to the tab that produced them and become
    stale after navigation or a major re-render.

  - `?list<string> documentIDs`

    References to files the harness has staged, for deployments where the browser executor cannot read the caller's filesystem.

  - `?list<string> paths`

    File paths on the browser executor's filesystem.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser File Upload Tool Use Block

- `class BrowserFileUploadToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserFileUploadInput input`

    Set the value of a file-input element to one or more files. The target must be an
    element reference; at least one of paths or document_ids is required.

  - `"file_upload" name`

  - `"browser" toolsetName`

### Browser Find Config

- `class BrowserFindConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Find Input

- `class BrowserFindInput`

  - `string query`

    Natural-language description of the element(s) to find.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Find Tool Use Block

- `class BrowserFindToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserFindInput input`

    Find elements matching a natural-language description (e.g. "search bar", "add to
    cart button") and return up to 20 matches with element references.

  - `"find" name`

  - `"browser" toolsetName`

### Browser Form Input Config

- `class BrowserFormInputConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Form Input Input

- `class BrowserFormInputInput`

  - `BrowserRefTarget target`

    An element on the page, identified by a reference from a prior `read_page` or
    `find` result. References are scoped to the tab that produced them and become
    stale after navigation or a major re-render.

  - `BrowserFormInputValue value`

    The value to set.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Form Input Tool Use Block

- `class BrowserFormInputToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserFormInputInput input`

    Set the value of a form element (input, textarea, select, checkbox). Use a
    boolean for checkboxes, an option value or text for selects.

  - `"form_input" name`

  - `"browser" toolsetName`

### Browser Form Input Value

- `class BrowserFormInputValue`

  - `string`

  - `float`

  - `bool`

### Browser Get Page Text Config

- `class BrowserGetPageTextConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Get Page Text Input

- `class BrowserGetPageTextInput`

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Get Page Text Tool Use Block

- `class BrowserGetPageTextToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserGetPageTextInput input`

    Return the page's visible text content as plain text, prioritizing article
    content. Suited to articles, documentation, and other text-heavy pages.

  - `"get_page_text" name`

  - `"browser" toolsetName`

### Browser Hold Key Config

- `class BrowserHoldKeyConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Hold Key Input

- `class BrowserHoldKeyInput`

  - `float duration`

    Seconds to hold the key down (maximum 30).

  - `string text`

    The key or chord to hold.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Hold Key Tool Use Block

- `class BrowserHoldKeyToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserHoldKeyInput input`

    Hold a key or key chord down for a duration, then release it. Uses the same key
    names and "+" chord syntax as the key action.

  - `"hold_key" name`

  - `"browser" toolsetName`

### Browser Hover Config

- `class BrowserHoverConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Hover Input

- `class BrowserHoverInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Hover Tool Use Block

- `class BrowserHoverToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserHoverInput input`

    Move the cursor to a coordinate or element without clicking.

  - `"hover" name`

  - `"browser" toolsetName`

### Browser Javascript Exec Config

- `class BrowserJavascriptExecConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Javascript Exec Input

- `class BrowserJavascriptExecInput`

  - `string text`

    JavaScript to execute in the page context.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Javascript Exec Tool Use Block

- `class BrowserJavascriptExecToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserJavascriptExecInput input`

    Execute JavaScript in the page context and return the value of the last
    expression. The code runs with access to the DOM, `window`, and page variables.
    Write the expression you want evaluated — do NOT use `return`.

  - `"javascript_exec" name`

  - `"browser" toolsetName`

### Browser Key Config

- `class BrowserKeyConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Key Input

- `class BrowserKeyInput`

  - `string text`

    The key, chord, or space-separated sequence to press.

  - `?int repeat`

    Number of times to repeat. Default 1.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Key Tool Use Block

- `class BrowserKeyToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserKeyInput input`

    Press a key or key chord. Use "+" to combine modifiers with a key (e.g. "ctrl+a",
    "cmd+shift+p") and space to sequence presses (e.g. "Backspace Backspace Delete").
    Common names like "Return", "Tab", "Escape", "BackSpace" are supported.

  - `"key" name`

  - `"browser" toolsetName`

### Browser Left Click Config

- `class BrowserLeftClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Left Click Drag Config

- `class BrowserLeftClickDragConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Left Click Drag Input

- `class BrowserLeftClickDragInput`

  - `BrowserCoordinateTarget from`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `BrowserCoordinateTarget target`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Left Click Drag Tool Use Block

- `class BrowserLeftClickDragToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserLeftClickDragInput input`

    Press at `from`, drag to `target`, release. Both must be coordinate targets.

  - `"left_click_drag" name`

  - `"browser" toolsetName`

### Browser Left Click Input

- `class BrowserLeftClickInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string modifiers`

    Optional modifier key chord to hold for the duration of this action (e.g. "shift", "ctrl+shift", "cmd+alt").

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Left Click Tool Use Block

- `class BrowserLeftClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserLeftClickInput input`

    Left-click at a viewport coordinate or on an element by reference.

  - `"left_click" name`

  - `"browser" toolsetName`

### Browser Left Mouse Down Config

- `class BrowserLeftMouseDownConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Left Mouse Down Input

- `class BrowserLeftMouseDownInput`

  - `BrowserCoordinateTarget target`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Left Mouse Down Tool Use Block

- `class BrowserLeftMouseDownToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserLeftMouseDownInput input`

    Press and hold the left mouse button at a viewport coordinate. Pair with
    left_mouse_up to perform a custom drag.

  - `"left_mouse_down" name`

  - `"browser" toolsetName`

### Browser Left Mouse Up Config

- `class BrowserLeftMouseUpConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Left Mouse Up Input

- `class BrowserLeftMouseUpInput`

  - `BrowserCoordinateTarget target`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Left Mouse Up Tool Use Block

- `class BrowserLeftMouseUpToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserLeftMouseUpInput input`

    Release the left mouse button at a viewport coordinate.

  - `"left_mouse_up" name`

  - `"browser" toolsetName`

### Browser List Tabs Config

- `class BrowserListTabsConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser List Tabs Input

- `BrowserListTabsInput`

### Browser List Tabs Tool Use Block

- `class BrowserListTabsToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserListTabsInput input`

    List all open tabs with each tab's tab_id, title, and URL.

  - `"list_tabs" name`

  - `"browser" toolsetName`

### Browser Middle Click Config

- `class BrowserMiddleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Middle Click Input

- `class BrowserMiddleClickInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string modifiers`

    Optional modifier key chord to hold for the duration of this action (e.g. "shift", "ctrl+shift", "cmd+alt").

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Middle Click Tool Use Block

- `class BrowserMiddleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserMiddleClickInput input`

    Middle-click at a viewport coordinate or on an element by reference.

  - `"middle_click" name`

  - `"browser" toolsetName`

### Browser Mouse Move Config

- `class BrowserMouseMoveConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Mouse Move Input

- `class BrowserMouseMoveInput`

  - `BrowserCoordinateTarget target`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Mouse Move Tool Use Block

- `class BrowserMouseMoveToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserMouseMoveInput input`

    Move the pointer to a viewport coordinate without clicking.

  - `"mouse_move" name`

  - `"browser" toolsetName`

### Browser Navigate Config

- `class BrowserNavigateConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Navigate Input

- `class BrowserNavigateInput`

  - `string url`

    The URL to navigate to, or "back" / "forward" / "reload" for history navigation.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Navigate Tool Use Block

- `class BrowserNavigateToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserNavigateInput input`

    Navigate to a URL, or go back/forward/reload in history. The protocol may be
    omitted (defaults to https://).

  - `"navigate" name`

  - `"browser" toolsetName`

### Browser New Tab Config

- `class BrowserNewTabConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser New Tab Input

- `BrowserNewTabInput`

### Browser New Tab Tool Use Block

- `class BrowserNewTabToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserNewTabInput input`

    Open a new empty tab and return its tab_id.

  - `"new_tab" name`

  - `"browser" toolsetName`

### Browser Read Console Config

- `class BrowserReadConsoleConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Read Console Input

- `class BrowserReadConsoleInput`

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Read Console Tool Use Block

- `class BrowserReadConsoleToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserReadConsoleInput input`

    Return console output (log entries, errors, warnings) accumulated since the
    driver attached to the tab and since the last read, one line per entry. An empty
    result does not mean no traffic for a tab that predates attach.

  - `"read_console" name`

  - `"browser" toolsetName`

### Browser Read Network Config

- `class BrowserReadNetworkConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Read Network Input

- `class BrowserReadNetworkInput`

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Read Network Tool Use Block

- `class BrowserReadNetworkToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserReadNetworkInput input`

    Return the network requests (method, URL, status, MIME type, timing) recorded
    since the driver attached to the tab and since the last read, one line per entry.
    An empty result does not mean no traffic for a tab that predates attach.

  - `"read_network" name`

  - `"browser" toolsetName`

### Browser Read Page Config

- `class BrowserReadPageConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Read Page Filter

- `enum BrowserReadPageFilter`

  - `"all"`

  - `"interactive"`

### Browser Read Page Input

- `class BrowserReadPageInput`

  - `?int depth`

    Maximum tree depth. Default 15.

  - `?BrowserReadPageFilter filter`

    Which elements to include. Omitted: every visible element. "interactive": interactive elements only. "all": additionally includes off-viewport elements.

  - `?string ref`

    Element reference to read a subtree from. Omit to read from the page root.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Read Page Tool Use Block

- `class BrowserReadPageToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserReadPageInput input`

    Return a structured accessibility tree of the page (or the subtree rooted at
    `ref`), with element references like [ref_7] that can be used as targets on later
    actions. Output is capped at 50,000 characters — narrow with `ref` or a smaller
    `depth` when exceeded.

  - `"read_page" name`

  - `"browser" toolsetName`

### Browser Ref Target

- `class BrowserRefTarget`

  - `"ref" type`

  - `string ref`

    An element reference (e.g. "ref_7") returned by a prior `read_page` or `find` result.

### Browser Right Click Config

- `class BrowserRightClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Right Click Input

- `class BrowserRightClickInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string modifiers`

    Optional modifier key chord to hold for the duration of this action (e.g. "shift", "ctrl+shift", "cmd+alt").

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Right Click Tool Use Block

- `class BrowserRightClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserRightClickInput input`

    Right-click at a viewport coordinate or on an element by reference.

  - `"right_click" name`

  - `"browser" toolsetName`

### Browser Screenshot Config

- `class BrowserScreenshotConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Screenshot Input

- `class BrowserScreenshotInput`

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Screenshot Tool Use Block

- `class BrowserScreenshotToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserScreenshotInput input`

    Capture the current browser viewport.

  - `"screenshot" name`

  - `"browser" toolsetName`

### Browser Scroll Config

- `class BrowserScrollConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Scroll Direction

- `enum BrowserScrollDirection`

  - `"up"`

  - `"down"`

  - `"left"`

  - `"right"`

### Browser Scroll Input

- `class BrowserScrollInput`

  - `BrowserScrollDirection scrollDirection`

  - `BrowserCoordinateTarget target`

    A point in the browser viewport, in viewport pixels (the same frame as a
    full-viewport screenshot).

  - `?int scrollAmount`

    Scroll-wheel notches (1–10). Default 3.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Scroll To Config

- `class BrowserScrollToConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Scroll To Input

- `class BrowserScrollToInput`

  - `BrowserRefTarget target`

    An element on the page, identified by a reference from a prior `read_page` or
    `find` result. References are scoped to the tab that produced them and become
    stale after navigation or a major re-render.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Scroll To Tool Use Block

- `class BrowserScrollToToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserScrollToInput input`

    Scroll an element into view.

  - `"scroll_to" name`

  - `"browser" toolsetName`

### Browser Scroll Tool Use Block

- `class BrowserScrollToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserScrollInput input`

    Scroll at a viewport position. `target` must be a coordinate target.

  - `"scroll" name`

  - `"browser" toolsetName`

### Browser State Block Param

- `class BrowserStateBlockParam`

  - `"browser_state" type`

  - `list<BrowserStateTabEntry> tabs`

    All tabs open in the browser after this call — the full inventory, not a delta. May be empty. Whenever non-empty, exactly one entry carries `active: true`.

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?list<BrowserStateChange> stateChanges`

    Tabs opened and download state changes during this call. "Nothing to report" is expressed by omitting the field, never by an empty list.

### Browser State Change

- `class BrowserStateChange`

  - `class BrowserStateChangeTabOpened`

    - `"tab_opened" type`

    - `string tabID`

      The `tab_id` of the opened tab, present in `tabs`.

  - `class BrowserStateChangeDownloadStarted`

    - `"download_started" type`

    - `string downloadID`

      The caller-assigned identifier for this download, stable across the state changes reporting it.

    - `string url`

      The final post-redirect URL the download was served from.

  - `class BrowserStateChangeDownloadCompleted`

    - `"download_completed" type`

    - `string downloadID`

      The caller-assigned identifier for this download, stable across the state changes reporting it.

    - `string url`

      The final post-redirect URL the download was served from.

    - `?string path`

      Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.

    - `?int sizeBytes`

      The completed download's size.

  - `class BrowserStateChangeDownloadFailed`

    - `"download_failed" type`

    - `string downloadID`

      The caller-assigned identifier for this download, stable across the state changes reporting it.

    - `string url`

      The final post-redirect URL the download was served from.

    - `?string error`

      The failure or cancellation detail, when known.

### Browser State Change Download Completed

- `class BrowserStateChangeDownloadCompleted`

  - `"download_completed" type`

  - `string downloadID`

    The caller-assigned identifier for this download, stable across the state changes reporting it.

  - `string url`

    The final post-redirect URL the download was served from.

  - `?string path`

    Where the executor saved the file, on the executor's filesystem. Only included when another tool in the same environment can read the file at that path.

  - `?int sizeBytes`

    The completed download's size.

### Browser State Change Download Failed

- `class BrowserStateChangeDownloadFailed`

  - `"download_failed" type`

  - `string downloadID`

    The caller-assigned identifier for this download, stable across the state changes reporting it.

  - `string url`

    The final post-redirect URL the download was served from.

  - `?string error`

    The failure or cancellation detail, when known.

### Browser State Change Download Started

- `class BrowserStateChangeDownloadStarted`

  - `"download_started" type`

  - `string downloadID`

    The caller-assigned identifier for this download, stable across the state changes reporting it.

  - `string url`

    The final post-redirect URL the download was served from.

### Browser State Change Tab Opened

- `class BrowserStateChangeTabOpened`

  - `"tab_opened" type`

  - `string tabID`

    The `tab_id` of the opened tab, present in `tabs`.

### Browser State Tab Entry

- `class BrowserStateTabEntry`

  - `string tabID`

    The caller-assigned identifier for this tab, unique within the inventory.

  - `string title`

    The title of the page the tab is showing. May be empty.

  - `string url`

    The URL of the page the tab is showing. May be empty.

  - `?bool active`

    Whether this tab is the active tab after this call. Whenever `tabs` is non-empty, exactly one entry is marked `active: true`.

### Browser Switch Tab Config

- `class BrowserSwitchTabConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Switch Tab Input

- `class BrowserSwitchTabInput`

  - `string tabID`

    The tab to switch to.

### Browser Switch Tab Tool Use Block

- `class BrowserSwitchTabToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserSwitchTabInput input`

    Make the tab with the given tab_id the active tab — the tab that actions without
    a tab_id apply to.

  - `"switch_tab" name`

  - `"browser" toolsetName`

### Browser Tool Use Block

- `class BrowserToolUseBlock`

  - `class BrowserNavigateToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserNavigateInput input`

      Navigate to a URL, or go back/forward/reload in history. The protocol may be
      omitted (defaults to https://).

    - `"navigate" name`

    - `"browser" toolsetName`

  - `class BrowserListTabsToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserListTabsInput input`

      List all open tabs with each tab's tab_id, title, and URL.

    - `"list_tabs" name`

    - `"browser" toolsetName`

  - `class BrowserNewTabToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserNewTabInput input`

      Open a new empty tab and return its tab_id.

    - `"new_tab" name`

    - `"browser" toolsetName`

  - `class BrowserSwitchTabToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserSwitchTabInput input`

      Make the tab with the given tab_id the active tab — the tab that actions without
      a tab_id apply to.

    - `"switch_tab" name`

    - `"browser" toolsetName`

  - `class BrowserCloseTabToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserCloseTabInput input`

      Close the tab with the given tab_id.

    - `"close_tab" name`

    - `"browser" toolsetName`

  - `class BrowserReadPageToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserReadPageInput input`

      Return a structured accessibility tree of the page (or the subtree rooted at
      `ref`), with element references like [ref_7] that can be used as targets on later
      actions. Output is capped at 50,000 characters — narrow with `ref` or a smaller
      `depth` when exceeded.

    - `"read_page" name`

    - `"browser" toolsetName`

  - `class BrowserGetPageTextToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserGetPageTextInput input`

      Return the page's visible text content as plain text, prioritizing article
      content. Suited to articles, documentation, and other text-heavy pages.

    - `"get_page_text" name`

    - `"browser" toolsetName`

  - `class BrowserReadConsoleToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserReadConsoleInput input`

      Return console output (log entries, errors, warnings) accumulated since the
      driver attached to the tab and since the last read, one line per entry. An empty
      result does not mean no traffic for a tab that predates attach.

    - `"read_console" name`

    - `"browser" toolsetName`

  - `class BrowserReadNetworkToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserReadNetworkInput input`

      Return the network requests (method, URL, status, MIME type, timing) recorded
      since the driver attached to the tab and since the last read, one line per entry.
      An empty result does not mean no traffic for a tab that predates attach.

    - `"read_network" name`

    - `"browser" toolsetName`

  - `class BrowserFindToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserFindInput input`

      Find elements matching a natural-language description (e.g. "search bar", "add to
      cart button") and return up to 20 matches with element references.

    - `"find" name`

    - `"browser" toolsetName`

  - `class BrowserFormInputToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserFormInputInput input`

      Set the value of a form element (input, textarea, select, checkbox). Use a
      boolean for checkboxes, an option value or text for selects.

    - `"form_input" name`

    - `"browser" toolsetName`

  - `class BrowserFileUploadToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserFileUploadInput input`

      Set the value of a file-input element to one or more files. The target must be an
      element reference; at least one of paths or document_ids is required.

    - `"file_upload" name`

    - `"browser" toolsetName`

  - `class BrowserScrollToToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserScrollToInput input`

      Scroll an element into view.

    - `"scroll_to" name`

    - `"browser" toolsetName`

  - `class BrowserScreenshotToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserScreenshotInput input`

      Capture the current browser viewport.

    - `"screenshot" name`

    - `"browser" toolsetName`

  - `class BrowserZoomToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserZoomInput input`

      Return a cropped screenshot of the given viewport region, scaled up for closer
      inspection — useful for small icons, buttons, or text. Coordinates are in the
      same viewport-pixel space as a full screenshot.

    - `"zoom" name`

    - `"browser" toolsetName`

  - `class BrowserLeftClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserLeftClickInput input`

      Left-click at a viewport coordinate or on an element by reference.

    - `"left_click" name`

    - `"browser" toolsetName`

  - `class BrowserRightClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserRightClickInput input`

      Right-click at a viewport coordinate or on an element by reference.

    - `"right_click" name`

    - `"browser" toolsetName`

  - `class BrowserMiddleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserMiddleClickInput input`

      Middle-click at a viewport coordinate or on an element by reference.

    - `"middle_click" name`

    - `"browser" toolsetName`

  - `class BrowserDoubleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserDoubleClickInput input`

      Double left-click at a viewport coordinate or on an element by reference.

    - `"double_click" name`

    - `"browser" toolsetName`

  - `class BrowserTripleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserTripleClickInput input`

      Triple left-click at a viewport coordinate or on an element by reference
      (typically selects a line or paragraph).

    - `"triple_click" name`

    - `"browser" toolsetName`

  - `class BrowserHoverToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserHoverInput input`

      Move the cursor to a coordinate or element without clicking.

    - `"hover" name`

    - `"browser" toolsetName`

  - `class BrowserLeftClickDragToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserLeftClickDragInput input`

      Press at `from`, drag to `target`, release. Both must be coordinate targets.

    - `"left_click_drag" name`

    - `"browser" toolsetName`

  - `class BrowserLeftMouseDownToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserLeftMouseDownInput input`

      Press and hold the left mouse button at a viewport coordinate. Pair with
      left_mouse_up to perform a custom drag.

    - `"left_mouse_down" name`

    - `"browser" toolsetName`

  - `class BrowserLeftMouseUpToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserLeftMouseUpInput input`

      Release the left mouse button at a viewport coordinate.

    - `"left_mouse_up" name`

    - `"browser" toolsetName`

  - `class BrowserMouseMoveToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserMouseMoveInput input`

      Move the pointer to a viewport coordinate without clicking.

    - `"mouse_move" name`

    - `"browser" toolsetName`

  - `class BrowserScrollToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserScrollInput input`

      Scroll at a viewport position. `target` must be a coordinate target.

    - `"scroll" name`

    - `"browser" toolsetName`

  - `class BrowserTypeToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserTypeInput input`

      Type a literal string at the current focus.

    - `"type" name`

    - `"browser" toolsetName`

  - `class BrowserKeyToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserKeyInput input`

      Press a key or key chord. Use "+" to combine modifiers with a key (e.g. "ctrl+a",
      "cmd+shift+p") and space to sequence presses (e.g. "Backspace Backspace Delete").
      Common names like "Return", "Tab", "Escape", "BackSpace" are supported.

    - `"key" name`

    - `"browser" toolsetName`

  - `class BrowserHoldKeyToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserHoldKeyInput input`

      Hold a key or key chord down for a duration, then release it. Uses the same key
      names and "+" chord syntax as the key action.

    - `"hold_key" name`

    - `"browser" toolsetName`

  - `class BrowserWaitToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserWaitInput input`

      Pause for the given duration.

    - `"wait" name`

    - `"browser" toolsetName`

  - `class BrowserJavascriptExecToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `BrowserJavascriptExecInput input`

      Execute JavaScript in the page context and return the value of the last
      expression. The code runs with access to the DOM, `window`, and page variables.
      Write the expression you want evaluated — do NOT use `return`.

    - `"javascript_exec" name`

    - `"browser" toolsetName`

### Browser Toolset 20260801

- `class BrowserToolset20260801`

  - `"browser_toolset_20260801" type`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?BrowserToolsetConfigs configs`

    Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

### Browser Toolset Configs

- `class BrowserToolsetConfigs`

  - `?BrowserTypeConfig type`

    `type`'s config overrides.

  - `?BrowserCloseTabConfig closeTab`

    `close_tab`'s config overrides.

  - `?BrowserDoubleClickConfig doubleClick`

    `double_click`'s config overrides.

  - `?BrowserFileUploadConfig fileUpload`

    `file_upload`'s config overrides.

  - `?BrowserFindConfig find`

    `find`'s config overrides.

  - `?BrowserFormInputConfig formInput`

    `form_input`'s config overrides.

  - `?BrowserGetPageTextConfig getPageText`

    `get_page_text`'s config overrides.

  - `?BrowserHoldKeyConfig holdKey`

    `hold_key`'s config overrides.

  - `?BrowserHoverConfig hover`

    `hover`'s config overrides.

  - `?BrowserJavascriptExecConfig javascriptExec`

    `javascript_exec`'s config overrides.

  - `?BrowserKeyConfig key`

    `key`'s config overrides.

  - `?BrowserLeftClickConfig leftClick`

    `left_click`'s config overrides.

  - `?BrowserLeftClickDragConfig leftClickDrag`

    `left_click_drag`'s config overrides.

  - `?BrowserLeftMouseDownConfig leftMouseDown`

    `left_mouse_down`'s config overrides.

  - `?BrowserLeftMouseUpConfig leftMouseUp`

    `left_mouse_up`'s config overrides.

  - `?BrowserListTabsConfig listTabs`

    `list_tabs`'s config overrides.

  - `?BrowserMiddleClickConfig middleClick`

    `middle_click`'s config overrides.

  - `?BrowserMouseMoveConfig mouseMove`

    `mouse_move`'s config overrides.

  - `?BrowserNavigateConfig navigate`

    `navigate`'s config overrides.

  - `?BrowserNewTabConfig newTab`

    `new_tab`'s config overrides.

  - `?BrowserReadConsoleConfig readConsole`

    `read_console`'s config overrides.

  - `?BrowserReadNetworkConfig readNetwork`

    `read_network`'s config overrides.

  - `?BrowserReadPageConfig readPage`

    `read_page`'s config overrides.

  - `?BrowserRightClickConfig rightClick`

    `right_click`'s config overrides.

  - `?BrowserScreenshotConfig screenshot`

    `screenshot`'s config overrides.

  - `?BrowserScrollConfig scroll`

    `scroll`'s config overrides.

  - `?BrowserScrollToConfig scrollTo`

    `scroll_to`'s config overrides.

  - `?BrowserSwitchTabConfig switchTab`

    `switch_tab`'s config overrides.

  - `?BrowserTripleClickConfig tripleClick`

    `triple_click`'s config overrides.

  - `?BrowserWaitConfig wait`

    `wait`'s config overrides.

  - `?BrowserZoomConfig zoom`

    `zoom`'s config overrides.

### Browser Triple Click Config

- `class BrowserTripleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Triple Click Input

- `class BrowserTripleClickInput`

  - `BrowserClickTarget target`

    Where to act: either a viewport coordinate or an element reference.

  - `?string modifiers`

    Optional modifier key chord to hold for the duration of this action (e.g. "shift", "ctrl+shift", "cmd+alt").

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Triple Click Tool Use Block

- `class BrowserTripleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserTripleClickInput input`

    Triple left-click at a viewport coordinate or on an element by reference
    (typically selects a line or paragraph).

  - `"triple_click" name`

  - `"browser" toolsetName`

### Browser Type Config

- `class BrowserTypeConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Type Input

- `class BrowserTypeInput`

  - `string text`

    The text to type.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Type Tool Use Block

- `class BrowserTypeToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserTypeInput input`

    Type a literal string at the current focus.

  - `"type" name`

  - `"browser" toolsetName`

### Browser Wait Config

- `class BrowserWaitConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Wait Input

- `class BrowserWaitInput`

  - `float duration`

    Seconds to wait (maximum 30).

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Wait Tool Use Block

- `class BrowserWaitToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserWaitInput input`

    Pause for the given duration.

  - `"wait" name`

  - `"browser" toolsetName`

### Browser Zoom Config

- `class BrowserZoomConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Browser Zoom Input

- `class BrowserZoomInput`

  - `list<int> region`

    [x0, y0, x1, y1] in viewport pixels.

  - `?string tabID`

    Tab to act on. Defaults to the active tab when omitted.

### Browser Zoom Tool Use Block

- `class BrowserZoomToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `BrowserZoomInput input`

    Return a cropped screenshot of the given viewport region, scaled up for closer
    inspection — useful for small icons, buttons, or text. Coordinates are in the
    same viewport-pixel space as a full screenshot.

  - `"zoom" name`

  - `"browser" toolsetName`

### Cache Control Ephemeral

- `class CacheControlEphemeral`

  - `"ephemeral" type`

  - `?TTL ttl`

    The time-to-live for the cache control breakpoint.

    This may be one the following values:

    - `5m`: 5 minutes
    - `1h`: 1 hour

    Defaults to `5m`. See [prompt caching pricing](https://platform.claude.com/docs/en/build-with-claude/prompt-caching?sdk_languages=php) for details.

### Cache Creation

- `class CacheCreation`

  - `int ephemeral1hInputTokens`

    The number of input tokens used to create the 1 hour cache entry.

  - `int ephemeral5mInputTokens`

    The number of input tokens used to create the 5 minute cache entry.

### Cache Miss Messages Changed

- `class CacheMissMessagesChanged`

  - `"messages_changed" type`

  - `int cacheMissedInputTokens`

    Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

### Cache Miss Model Changed

- `class CacheMissModelChanged`

  - `"model_changed" type`

  - `int cacheMissedInputTokens`

    Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

### Cache Miss Previous Message Not Found

- `class CacheMissPreviousMessageNotFound`

  - `"previous_message_not_found" type`

### Cache Miss Reason

- `class CacheMissReason`

  - `class CacheMissModelChanged`

    - `"model_changed" type`

    - `int cacheMissedInputTokens`

      Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

  - `class CacheMissSystemChanged`

    - `"system_changed" type`

    - `int cacheMissedInputTokens`

      Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

  - `class CacheMissToolsChanged`

    - `"tools_changed" type`

    - `int cacheMissedInputTokens`

      Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

  - `class CacheMissMessagesChanged`

    - `"messages_changed" type`

    - `int cacheMissedInputTokens`

      Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

  - `class CacheMissPreviousMessageNotFound`

    - `"previous_message_not_found" type`

  - `class CacheMissUnavailable`

    - `"unavailable" type`

### Cache Miss System Changed

- `class CacheMissSystemChanged`

  - `"system_changed" type`

  - `int cacheMissedInputTokens`

    Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

### Cache Miss Tools Changed

- `class CacheMissToolsChanged`

  - `"tools_changed" type`

  - `int cacheMissedInputTokens`

    Approximate number of input tokens that would have been read from cache had the prefix matched the previous request.

### Cache Miss Unavailable

- `class CacheMissUnavailable`

  - `"unavailable" type`

### Citation Char Location

- `class CitationCharLocation`

  - `"char_location" type`

  - `string citedText`

  - `int documentIndex`

  - `?string documentTitle`

  - `int endCharIndex`

  - `?string fileID`

  - `int startCharIndex`

### Citation Char Location Param

- `class CitationCharLocationParam`

  - `"char_location" type`

  - `string citedText`

  - `int documentIndex`

  - `?string documentTitle`

  - `int endCharIndex`

  - `int startCharIndex`

### Citation Content Block Location

- `class CitationContentBlockLocation`

  - `"content_block_location" type`

  - `string citedText`

    The full text of the cited block range, concatenated.

    Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

  - `int documentIndex`

  - `?string documentTitle`

  - `int endBlockIndex`

    Exclusive 0-based end index of the cited block range in the source's `content` array.

    Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

  - `?string fileID`

  - `int startBlockIndex`

    0-based index of the first cited block in the source's `content` array.

### Citation Content Block Location Param

- `class CitationContentBlockLocationParam`

  - `"content_block_location" type`

  - `string citedText`

    The full text of the cited block range, concatenated.

    Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

  - `int documentIndex`

  - `?string documentTitle`

  - `int endBlockIndex`

    Exclusive 0-based end index of the cited block range in the source's `content` array.

    Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

  - `int startBlockIndex`

    0-based index of the first cited block in the source's `content` array.

### Citation Page Location

- `class CitationPageLocation`

  - `"page_location" type`

  - `string citedText`

  - `int documentIndex`

  - `?string documentTitle`

  - `int endPageNumber`

  - `?string fileID`

  - `int startPageNumber`

### Citation Page Location Param

- `class CitationPageLocationParam`

  - `"page_location" type`

  - `string citedText`

  - `int documentIndex`

  - `?string documentTitle`

  - `int endPageNumber`

  - `int startPageNumber`

### Citation Search Result Location Param

- `class CitationSearchResultLocationParam`

  - `"search_result_location" type`

  - `string citedText`

    The full text of the cited block range, concatenated.

    Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

  - `int endBlockIndex`

    Exclusive 0-based end index of the cited block range in the source's `content` array.

    Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

  - `int searchResultIndex`

    0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.

    Counted separately from `document_index`; server-side web search results are not included in this count.

  - `string source`

  - `int startBlockIndex`

    0-based index of the first cited block in the source's `content` array.

  - `?string title`

### Citation Web Search Result Location Param

- `class CitationWebSearchResultLocationParam`

  - `"web_search_result_location" type`

  - `string citedText`

  - `string encryptedIndex`

  - `?string title`

  - `string url`

### Citations Config

- `class CitationsConfig`

  - `bool enabled`

### Citations Config Param

- `class CitationsConfigParam`

  - `?bool enabled`

### Citations Delta

- `class CitationsDelta`

  - `"citations_delta" type`

  - `Citation citation`

### Citations Search Result Location

- `class CitationsSearchResultLocation`

  - `"search_result_location" type`

  - `string citedText`

    The full text of the cited block range, concatenated.

    Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

  - `int endBlockIndex`

    Exclusive 0-based end index of the cited block range in the source's `content` array.

    Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

  - `int searchResultIndex`

    0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.

    Counted separately from `document_index`; server-side web search results are not included in this count.

  - `string source`

  - `int startBlockIndex`

    0-based index of the first cited block in the source's `content` array.

  - `?string title`

### Citations Web Search Result Location

- `class CitationsWebSearchResultLocation`

  - `"web_search_result_location" type`

  - `string citedText`

  - `string encryptedIndex`

  - `?string title`

  - `string url`

### Code Execution Output Block

- `class CodeExecutionOutputBlock`

  - `"code_execution_output" type`

  - `string fileID`

### Code Execution Output Block Param

- `class CodeExecutionOutputBlockParam`

  - `"code_execution_output" type`

  - `string fileID`

### Code Execution Result Block

- `class CodeExecutionResultBlock`

  - `"code_execution_result" type`

  - `list<CodeExecutionOutputBlock> content`

  - `int returnCode`

  - `string stderr`

  - `string stdout`

### Code Execution Result Block Param

- `class CodeExecutionResultBlockParam`

  - `"code_execution_result" type`

  - `list<CodeExecutionOutputBlockParam> content`

  - `int returnCode`

  - `string stderr`

  - `string stdout`

### Code Execution Tool 20250522

- `class CodeExecutionTool20250522`

  - `"code_execution_20250522" type`

  - `"code_execution" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Code Execution Tool 20250825

- `class CodeExecutionTool20250825`

  - `"code_execution_20250825" type`

  - `"code_execution" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Code Execution Tool 20260120

- `class CodeExecutionTool20260120`

  - `"code_execution_20260120" type`

  - `"code_execution" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Code Execution Tool 20260521

- `class CodeExecutionTool20260521`

  - `"code_execution_20260521" type`

  - `"code_execution" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Code Execution Tool Result Block

- `class CodeExecutionToolResultBlock`

  - `"code_execution_tool_result" type`

  - `CodeExecutionToolResultBlockContent content`

  - `string toolUseID`

### Code Execution Tool Result Block Content

- `class CodeExecutionToolResultBlockContent`

  - `class CodeExecutionToolResultError`

    - `"code_execution_tool_result_error" type`

    - `CodeExecutionToolResultErrorCode errorCode`

  - `class CodeExecutionResultBlock`

    - `"code_execution_result" type`

    - `list<CodeExecutionOutputBlock> content`

    - `int returnCode`

    - `string stderr`

    - `string stdout`

  - `class EncryptedCodeExecutionResultBlock`

    - `"encrypted_code_execution_result" type`

    - `list<CodeExecutionOutputBlock> content`

    - `string encryptedStdout`

    - `int returnCode`

    - `string stderr`

### Code Execution Tool Result Block Param

- `class CodeExecutionToolResultBlockParam`

  - `"code_execution_tool_result" type`

  - `CodeExecutionToolResultBlockParamContent content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Code Execution Tool Result Block Param Content

- `class CodeExecutionToolResultBlockParamContent`

  - `class CodeExecutionToolResultErrorParam`

    - `"code_execution_tool_result_error" type`

    - `CodeExecutionToolResultErrorCode errorCode`

  - `class CodeExecutionResultBlockParam`

    - `"code_execution_result" type`

    - `list<CodeExecutionOutputBlockParam> content`

    - `int returnCode`

    - `string stderr`

    - `string stdout`

  - `class EncryptedCodeExecutionResultBlockParam`

    - `"encrypted_code_execution_result" type`

    - `list<CodeExecutionOutputBlockParam> content`

    - `string encryptedStdout`

    - `int returnCode`

    - `string stderr`

### Code Execution Tool Result Error

- `class CodeExecutionToolResultError`

  - `"code_execution_tool_result_error" type`

  - `CodeExecutionToolResultErrorCode errorCode`

### Code Execution Tool Result Error Code

- `enum CodeExecutionToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"unavailable"`

  - `"too_many_requests"`

  - `"execution_time_exceeded"`

### Code Execution Tool Result Error Param

- `class CodeExecutionToolResultErrorParam`

  - `"code_execution_tool_result_error" type`

  - `CodeExecutionToolResultErrorCode errorCode`

### Computer Cursor Position Config

- `class ComputerCursorPositionConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Cursor Position Input

- `ComputerCursorPositionInput`

### Computer Cursor Position Tool Use Block

- `class ComputerCursorPositionToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerCursorPositionInput input`

    Get the current (x, y) pixel coordinate of the cursor.

  - `"cursor_position" name`

  - `"computer" toolsetName`

### Computer Double Click Config

- `class ComputerDoubleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Double Click Input

- `class ComputerDoubleClickInput`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Double Click Tool Use Block

- `class ComputerDoubleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerDoubleClickInput input`

    Double-click the left mouse button at the specified (x, y) pixel coordinate, or
    the current cursor position if `coordinate` is omitted.

  - `"double_click" name`

  - `"computer" toolsetName`

### Computer Hold Key Config

- `class ComputerHoldKeyConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Hold Key Input

- `class ComputerHoldKeyInput`

  - `int duration`

    Duration to hold the key, in seconds.

  - `string text`

    The key or key-combination to hold.

### Computer Hold Key Tool Use Block

- `class ComputerHoldKeyToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerHoldKeyInput input`

    Hold down a key or key-combination for a specified duration. Uses the same key
    syntax as `key`.

  - `"hold_key" name`

  - `"computer" toolsetName`

### Computer Key Config

- `class ComputerKeyConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Key Input

- `class ComputerKeyInput`

  - `string text`

    The key or key-combination to press.

  - `?int repeat`

    Number of times to repeat the key press. Default is 1.

### Computer Key Tool Use Block

- `class ComputerKeyToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerKeyInput input`

    Press a key or key-combination on the keyboard. Use "+" to combine modifiers with
    a key (e.g. "ctrl+s", "alt+Tab", "ctrl+shift+Escape"). Key names are
    case-insensitive; common names like "Return", "Tab", "Escape", "Up", "Down",
    "Left", "Right", "Home", "End", "Page_Up", "Page_Down", "Delete", "BackSpace" are
    supported.

  - `"key" name`

  - `"computer" toolsetName`

### Computer Left Click Config

- `class ComputerLeftClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Left Click Drag Config

- `class ComputerLeftClickDragConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Left Click Drag Input

- `class ComputerLeftClickDragInput`

  - `list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `list<int> startCoordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Left Click Drag Tool Use Block

- `class ComputerLeftClickDragToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerLeftClickDragInput input`

    Click and drag the cursor from `start_coordinate` to `coordinate`.

  - `"left_click_drag" name`

  - `"computer" toolsetName`

### Computer Left Click Input

- `class ComputerLeftClickInput`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Left Click Tool Use Block

- `class ComputerLeftClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerLeftClickInput input`

    Click the left mouse button at the specified (x, y) pixel coordinate, or the
    current cursor position if `coordinate` is omitted.

  - `"left_click" name`

  - `"computer" toolsetName`

### Computer Left Mouse Down Config

- `class ComputerLeftMouseDownConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Left Mouse Down Input

- `ComputerLeftMouseDownInput`

### Computer Left Mouse Down Tool Use Block

- `class ComputerLeftMouseDownToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerLeftMouseDownInput input`

    Press and hold the left mouse button at the current cursor position.

  - `"left_mouse_down" name`

  - `"computer" toolsetName`

### Computer Left Mouse Up Config

- `class ComputerLeftMouseUpConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Left Mouse Up Input

- `ComputerLeftMouseUpInput`

### Computer Left Mouse Up Tool Use Block

- `class ComputerLeftMouseUpToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerLeftMouseUpInput input`

    Release the left mouse button.

  - `"left_mouse_up" name`

  - `"computer" toolsetName`

### Computer Middle Click Config

- `class ComputerMiddleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Middle Click Input

- `class ComputerMiddleClickInput`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Middle Click Tool Use Block

- `class ComputerMiddleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerMiddleClickInput input`

    Click the middle mouse button at the specified (x, y) pixel coordinate, or the
    current cursor position if `coordinate` is omitted.

  - `"middle_click" name`

  - `"computer" toolsetName`

### Computer Mouse Move Config

- `class ComputerMouseMoveConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Mouse Move Input

- `class ComputerMouseMoveInput`

  - `list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

### Computer Mouse Move Tool Use Block

- `class ComputerMouseMoveToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerMouseMoveInput input`

    Move the cursor to a specified (x, y) pixel coordinate. Use this ONLY to hover
    without clicking; otherwise use a click action directly.

  - `"mouse_move" name`

  - `"computer" toolsetName`

### Computer Right Click Config

- `class ComputerRightClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Right Click Input

- `class ComputerRightClickInput`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Right Click Tool Use Block

- `class ComputerRightClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerRightClickInput input`

    Click the right mouse button at the specified (x, y) pixel coordinate, or the
    current cursor position if `coordinate` is omitted.

  - `"right_click" name`

  - `"computer" toolsetName`

### Computer Screenshot Config

- `class ComputerScreenshotConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Screenshot Input

- `ComputerScreenshotInput`

### Computer Screenshot Tool Use Block

- `class ComputerScreenshotToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerScreenshotInput input`

    Take a screenshot of the screen.

  - `"screenshot" name`

  - `"computer" toolsetName`

### Computer Scroll Config

- `class ComputerScrollConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Scroll Direction

- `enum ComputerScrollDirection`

  - `"up"`

  - `"down"`

  - `"left"`

  - `"right"`

### Computer Scroll Input

- `class ComputerScrollInput`

  - `int scrollAmount`

    Number of 'clicks' of the scroll wheel.

  - `ComputerScrollDirection scrollDirection`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Scroll Tool Use Block

- `class ComputerScrollToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerScrollInput input`

    Scroll the screen at the specified (x, y) pixel coordinate, or the current cursor
    position if `coordinate` is omitted. Do NOT use PageUp/PageDown to scroll.

  - `"scroll" name`

  - `"computer" toolsetName`

### Computer Tool Use Block

- `class ComputerToolUseBlock`

  - `class ComputerKeyToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerKeyInput input`

      Press a key or key-combination on the keyboard. Use "+" to combine modifiers with
      a key (e.g. "ctrl+s", "alt+Tab", "ctrl+shift+Escape"). Key names are
      case-insensitive; common names like "Return", "Tab", "Escape", "Up", "Down",
      "Left", "Right", "Home", "End", "Page_Up", "Page_Down", "Delete", "BackSpace" are
      supported.

    - `"key" name`

    - `"computer" toolsetName`

  - `class ComputerHoldKeyToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerHoldKeyInput input`

      Hold down a key or key-combination for a specified duration. Uses the same key
      syntax as `key`.

    - `"hold_key" name`

    - `"computer" toolsetName`

  - `class ComputerTypeToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerTypeInput input`

      Type a string of text on the keyboard.

    - `"type" name`

    - `"computer" toolsetName`

  - `class ComputerCursorPositionToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerCursorPositionInput input`

      Get the current (x, y) pixel coordinate of the cursor.

    - `"cursor_position" name`

    - `"computer" toolsetName`

  - `class ComputerMouseMoveToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerMouseMoveInput input`

      Move the cursor to a specified (x, y) pixel coordinate. Use this ONLY to hover
      without clicking; otherwise use a click action directly.

    - `"mouse_move" name`

    - `"computer" toolsetName`

  - `class ComputerLeftMouseDownToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerLeftMouseDownInput input`

      Press and hold the left mouse button at the current cursor position.

    - `"left_mouse_down" name`

    - `"computer" toolsetName`

  - `class ComputerLeftMouseUpToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerLeftMouseUpInput input`

      Release the left mouse button.

    - `"left_mouse_up" name`

    - `"computer" toolsetName`

  - `class ComputerLeftClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerLeftClickInput input`

      Click the left mouse button at the specified (x, y) pixel coordinate, or the
      current cursor position if `coordinate` is omitted.

    - `"left_click" name`

    - `"computer" toolsetName`

  - `class ComputerLeftClickDragToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerLeftClickDragInput input`

      Click and drag the cursor from `start_coordinate` to `coordinate`.

    - `"left_click_drag" name`

    - `"computer" toolsetName`

  - `class ComputerRightClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerRightClickInput input`

      Click the right mouse button at the specified (x, y) pixel coordinate, or the
      current cursor position if `coordinate` is omitted.

    - `"right_click" name`

    - `"computer" toolsetName`

  - `class ComputerMiddleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerMiddleClickInput input`

      Click the middle mouse button at the specified (x, y) pixel coordinate, or the
      current cursor position if `coordinate` is omitted.

    - `"middle_click" name`

    - `"computer" toolsetName`

  - `class ComputerDoubleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerDoubleClickInput input`

      Double-click the left mouse button at the specified (x, y) pixel coordinate, or
      the current cursor position if `coordinate` is omitted.

    - `"double_click" name`

    - `"computer" toolsetName`

  - `class ComputerTripleClickToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerTripleClickInput input`

      Triple-click the left mouse button at the specified (x, y) pixel coordinate, or
      the current cursor position if `coordinate` is omitted.

    - `"triple_click" name`

    - `"computer" toolsetName`

  - `class ComputerScrollToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerScrollInput input`

      Scroll the screen at the specified (x, y) pixel coordinate, or the current cursor
      position if `coordinate` is omitted. Do NOT use PageUp/PageDown to scroll.

    - `"scroll" name`

    - `"computer" toolsetName`

  - `class ComputerWaitToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerWaitInput input`

      Wait for a specified duration.

    - `"wait" name`

    - `"computer" toolsetName`

  - `class ComputerScreenshotToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerScreenshotInput input`

      Take a screenshot of the screen.

    - `"screenshot" name`

    - `"computer" toolsetName`

  - `class ComputerZoomToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `ToolUseCaller caller`

      Which party invoked the tool call: the model directly, or a server tool on its behalf.

    - `ComputerZoomInput input`

      Take a screenshot of a rectangular region. Region coordinates are in the
      full-screenshot space (not physical display pixels). The crop is scaled up to
      fill the image budget so fine details become legible.

    - `"zoom" name`

    - `"computer" toolsetName`

### Computer Toolset 20260801

- `class ComputerToolset20260801`

  - `"computer_toolset_20260801" type`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?ComputerToolsetConfigs configs`

    Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

### Computer Toolset Configs

- `class ComputerToolsetConfigs`

  - `?ComputerTypeConfig type`

    `type`'s config overrides.

  - `?ComputerCursorPositionConfig cursorPosition`

    `cursor_position`'s config overrides.

  - `?ComputerDoubleClickConfig doubleClick`

    `double_click`'s config overrides.

  - `?ComputerHoldKeyConfig holdKey`

    `hold_key`'s config overrides.

  - `?ComputerKeyConfig key`

    `key`'s config overrides.

  - `?ComputerLeftClickConfig leftClick`

    `left_click`'s config overrides.

  - `?ComputerLeftClickDragConfig leftClickDrag`

    `left_click_drag`'s config overrides.

  - `?ComputerLeftMouseDownConfig leftMouseDown`

    `left_mouse_down`'s config overrides.

  - `?ComputerLeftMouseUpConfig leftMouseUp`

    `left_mouse_up`'s config overrides.

  - `?ComputerMiddleClickConfig middleClick`

    `middle_click`'s config overrides.

  - `?ComputerMouseMoveConfig mouseMove`

    `mouse_move`'s config overrides.

  - `?ComputerRightClickConfig rightClick`

    `right_click`'s config overrides.

  - `?ComputerScreenshotConfig screenshot`

    `screenshot`'s config overrides.

  - `?ComputerScrollConfig scroll`

    `scroll`'s config overrides.

  - `?ComputerTripleClickConfig tripleClick`

    `triple_click`'s config overrides.

  - `?ComputerWaitConfig wait`

    `wait`'s config overrides.

  - `?ComputerZoomConfig zoom`

    `zoom`'s config overrides.

### Computer Triple Click Config

- `class ComputerTripleClickConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Triple Click Input

- `class ComputerTripleClickInput`

  - `?list<int> coordinate`

    (x, y): x pixels from the left edge, y pixels from the top edge.

  - `?string text`

    Optional key combination to hold down during this action (e.g. "ctrl", "shift", "ctrl+shift").

### Computer Triple Click Tool Use Block

- `class ComputerTripleClickToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerTripleClickInput input`

    Triple-click the left mouse button at the specified (x, y) pixel coordinate, or
    the current cursor position if `coordinate` is omitted.

  - `"triple_click" name`

  - `"computer" toolsetName`

### Computer Type Config

- `class ComputerTypeConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Type Input

- `class ComputerTypeInput`

  - `string text`

    The text to type.

### Computer Type Tool Use Block

- `class ComputerTypeToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerTypeInput input`

    Type a string of text on the keyboard.

  - `"type" name`

  - `"computer" toolsetName`

### Computer Wait Config

- `class ComputerWaitConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Wait Input

- `class ComputerWaitInput`

  - `int duration`

    Duration to wait, in seconds.

### Computer Wait Tool Use Block

- `class ComputerWaitToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerWaitInput input`

    Wait for a specified duration.

  - `"wait" name`

  - `"computer" toolsetName`

### Computer Zoom Config

- `class ComputerZoomConfig`

  - `?bool deferLoading`

    Defer loading for this member. Must resolve to the same value on every enabled member of the toolset.

  - `?bool enabled`

    Whether this member is offered to the model. Default is per member, per the toolset's documentation. A member whose enabled resolves false is withheld from the served schema.

### Computer Zoom Input

- `class ComputerZoomInput`

  - `list<int> region`

    (x0, y0, x1, y1): The region to capture.

### Computer Zoom Tool Use Block

- `class ComputerZoomToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `ToolUseCaller caller`

    Which party invoked the tool call: the model directly, or a server tool on its behalf.

  - `ComputerZoomInput input`

    Take a screenshot of a rectangular region. Region coordinates are in the
    full-screenshot space (not physical display pixels). The crop is scaled up to
    fill the image budget so fine details become legible.

  - `"zoom" name`

  - `"computer" toolsetName`

### Container

- `class Container`

  - `string id`

    Identifier for the container used in this request

  - `\Datetime expiresAt`

    The time at which the container will expire.

  - `?list<ContainerSkill> skills`

    Skills loaded in the container

### Container Params

- `class ContainerParams`

  - `?string id`

    Container id

  - `?list<SkillParams> skills`

    List of skills to load in the container

### Container Skill

- `class ContainerSkill`

  - `Type type`

    Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)

  - `string skillID`

    Skill ID

  - `string version`

    The resolved version: a skill version ID for custom skills.

### Container Upload Block

- `class ContainerUploadBlock`

  - `"container_upload" type`

  - `string fileID`

### Container Upload Block Param

- `class ContainerUploadBlockParam`

  - `"container_upload" type`

  - `string fileID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Content Block

- `class ContentBlock`

  - `class TextBlock`

    - `"text" type`

    - `?list<TextCitation> citations`

      Citations supporting the text block.

      The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.

    - `string text`

  - `class ThinkingBlock`

    - `"thinking" type`

    - `string signature`

      A value used to verify that this thinking block was generated by Claude when it is passed back to the API.

      This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.

      See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php) for details.

    - `string thinking`

      The text of Claude's thinking process for this block.

  - `class RedactedThinkingBlock`

    - `"redacted_thinking" type`

    - `string data`

      The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.

      Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.

      See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php#redacted-thinking-blocks) for details.

  - `class ToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `Caller caller`

    - `array<string,mixed> input`

    - `string name`

    - `?string toolsetName`

      For a toolset member tool_use, the toolset family.

  - `class ServerToolUseBlock`

    - `"server_tool_use" type`

    - `string id`

    - `Caller caller`

    - `array<string,mixed> input`

    - `Name name`

  - `class WebSearchToolResultBlock`

    - `"web_search_tool_result" type`

    - `Caller caller`

    - `WebSearchToolResultBlockContent content`

    - `string toolUseID`

  - `class WebFetchToolResultBlock`

    - `"web_fetch_tool_result" type`

    - `Caller caller`

    - `Content content`

    - `string toolUseID`

  - `class CodeExecutionToolResultBlock`

    - `"code_execution_tool_result" type`

    - `CodeExecutionToolResultBlockContent content`

    - `string toolUseID`

  - `class BashCodeExecutionToolResultBlock`

    - `"bash_code_execution_tool_result" type`

    - `Content content`

    - `string toolUseID`

  - `class TextEditorCodeExecutionToolResultBlock`

    - `"text_editor_code_execution_tool_result" type`

    - `Content content`

    - `string toolUseID`

  - `class ToolSearchToolResultBlock`

    - `"tool_search_tool_result" type`

    - `Content content`

    - `string toolUseID`

  - `class ContainerUploadBlock`

    - `"container_upload" type`

    - `string fileID`

### Content Block Param

- `class ContentBlockParam`

  - `class TextBlockParam`

    - `"text" type`

    - `string text`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?list<TextCitationParam> citations`

  - `class ImageBlockParam`

    - `"image" type`

    - `Source source`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?ImageTransformationsParam transformations`

      Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.

  - `class DocumentBlockParam`

    - `"document" type`

    - `Source source`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

    - `?string context`

    - `?string title`

  - `class SearchResultBlockParam`

    - `"search_result" type`

    - `list<TextBlockParam> content`

    - `string source`

    - `string title`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

  - `class ThinkingBlockParam`

    - `"thinking" type`

    - `string signature`

      The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.

      Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.

    - `string thinking`

      The `thinking` text of this block as returned by the API.

  - `class RedactedThinkingBlockParam`

    - `"redacted_thinking" type`

    - `string data`

      The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.

  - `class ToolUseBlockParam`

    - `"tool_use" type`

    - `string id`

    - `array<string,mixed> input`

    - `string name`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?Caller caller`

    - `?string toolsetName`

      For a toolset member tool_use, the toolset family this member belongs to.

  - `class ToolResultBlockParam`

    - `"tool_result" type`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?Content content`

    - `?bool isError`

    - `?string toolsetName`

      For a toolset member tool_result, the toolset family of the paired tool_use.

  - `class ServerToolUseBlockParam`

    - `"server_tool_use" type`

    - `string id`

    - `array<string,mixed> input`

    - `Name name`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?Caller caller`

  - `class WebSearchToolResultBlockParam`

    - `"web_search_tool_result" type`

    - `WebSearchToolResultBlockParamContent content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?Caller caller`

  - `class WebFetchToolResultBlockParam`

    - `"web_fetch_tool_result" type`

    - `Content content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?Caller caller`

  - `class CodeExecutionToolResultBlockParam`

    - `"code_execution_tool_result" type`

    - `CodeExecutionToolResultBlockParamContent content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

  - `class BashCodeExecutionToolResultBlockParam`

    - `"bash_code_execution_tool_result" type`

    - `Content content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

  - `class TextEditorCodeExecutionToolResultBlockParam`

    - `"text_editor_code_execution_tool_result" type`

    - `Content content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

  - `class ToolSearchToolResultBlockParam`

    - `"tool_search_tool_result" type`

    - `Content content`

    - `string toolUseID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

  - `class ContainerUploadBlockParam`

    - `"container_upload" type`

    - `string fileID`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

### Content Block Source

- `class ContentBlockSource`

  - `"content" type`

  - `Content content`

### Content Block Source Content

- `class ContentBlockSourceContent`

  - `class TextBlockParam`

    - `"text" type`

    - `string text`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?list<TextCitationParam> citations`

  - `class ImageBlockParam`

    - `"image" type`

    - `Source source`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?ImageTransformationsParam transformations`

      Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.

### Diagnostics

- `class Diagnostics`

  - `?CacheMissReason cacheMissReason`

    Explains why the prompt cache could not fully reuse the prefix from the request identified by `diagnostics.previous_message_id`. `null` means diagnosis is still pending — the response was serialized before the background comparison completed.

### Diagnostics Param

- `class DiagnosticsParam`

  - `?string previousMessageID`

    The `id` (`msg_...`) from this client's previous /v1/messages response. The server compares that request's prompt fingerprint against this one and returns `diagnostics.cache_miss_reason` when the prompt-cache prefix could not be reused. Pass `null` on the first turn to opt in without a prior message to compare.

### Direct Caller

- `class DirectCaller`

  - `"direct" type`

### Document Block

- `class DocumentBlock`

  - `"document" type`

  - `?CitationsConfig citations`

    Citation configuration for the document

  - `Source source`

  - `?string title`

    The title of the document

### Document Block Param

- `class DocumentBlockParam`

  - `"document" type`

  - `Source source`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

  - `?string context`

  - `?string title`

### Encrypted Code Execution Result Block

- `class EncryptedCodeExecutionResultBlock`

  - `"encrypted_code_execution_result" type`

  - `list<CodeExecutionOutputBlock> content`

  - `string encryptedStdout`

  - `int returnCode`

  - `string stderr`

### Encrypted Code Execution Result Block Param

- `class EncryptedCodeExecutionResultBlockParam`

  - `"encrypted_code_execution_result" type`

  - `list<CodeExecutionOutputBlockParam> content`

  - `string encryptedStdout`

  - `int returnCode`

  - `string stderr`

### File Document Source

- `class FileDocumentSource`

  - `"file" type`

  - `string fileID`

### File Image Source

- `class FileImageSource`

  - `"file" type`

  - `string fileID`

### Image Block Param

- `class ImageBlockParam`

  - `"image" type`

  - `Source source`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?ImageTransformationsParam transformations`

    Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.

### Image Transformations Param

- `class ImageTransformationsParam`

  - `?OversizedImage oversizedImage`

    What the server does when this image exceeds the model's maximum image size. `"downsize"` (the default) scales the image down to fit, which changes the dimensions the model observes without telling you. `"error"` instead rejects the request with a 400 error naming the image's dimensions and the largest dimensions that fit, so you can scale the image deliberately — your image is never silently scaled down.

### Input JSON Delta

- `class InputJSONDelta`

  - `"input_json_delta" type`

  - `string partialJSON`

### JSON Output Format

- `class JSONOutputFormat`

  - `"json_schema" type`

  - `array<string,mixed> schema`

    The JSON schema of the format

### Memory Tool 20250818

- `class MemoryTool20250818`

  - `"memory_20250818" type`

  - `"memory" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?list<array<string,mixed>> inputExamples`

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Message

- `class Message`

  - `"message" type`

    Object type.

    For Messages, this is always `"message"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?Container container`

    Information about the container used in this request.

    This will be non-null if a container tool (e.g. code execution) was used.

  - `list<ContentBlock> content`

    Content generated by the model.

    This is an array of content blocks, each of which has a `type` that determines its shape.

    Example:

    ```json
    [{"type": "text", "text": "Hi, I'm Claude."}]
    ```

    If the request input `messages` ended with an `assistant` turn, then the response `content` will continue directly from that last turn. You can use this to constrain the model's output.

    For example, if the input `messages` were:

    ```json
    [
      {"role": "user", "content": "What's the Greek name for Sun? (A) Sol (B) Helios (C) Sun"},
      {"role": "assistant", "content": "The best answer is ("}
    ]
    ```

    Then the response `content` might be:

    ```json
    [{"type": "text", "text": "B)"}]
    ```

  - `?Diagnostics diagnostics`

    Request-level diagnostics. `null` when the request did not supply `diagnostics`, or when it did and no prompt-cache divergence was detected.

  - `Model model`

    The model that will complete your prompt.

    See [models](https://docs.anthropic.com/en/docs/models-overview) for additional details and options.

  - `"assistant" role`

    Conversational role of the generated message.

    This will always be `"assistant"`.

  - `?RefusalStopDetails stopDetails`

    Structured information about why model output stopped.

    This is `null` when the `stop_reason` has no additional detail to report.

  - `?StopReason stopReason`

    The reason that we stopped.

    This may be one the following values:

    * `"end_turn"`: the model reached a natural stopping point
    * `"max_tokens"`: we exceeded the requested `max_tokens` or the model's maximum
    * `"stop_sequence"`: one of your provided custom `stop_sequences` was generated
    * `"tool_use"`: the model invoked one or more tools
    * `"pause_turn"`: we paused a long-running turn. You may provide the response back as-is in a subsequent request to let the model continue.
    * `"refusal"`: when streaming classifiers intervene to handle potential policy violations
    * `"model_context_window_exceeded"`: we exceeded the model's context window

    In non-streaming mode this value is always non-null. In streaming mode, it is null in the `message_start` event and non-null otherwise.

  - `?string stopSequence`

    Which custom stop sequence was generated, if any.

    This value will be a non-null string if one of your custom stop sequences was generated.

  - `Usage usage`

    Billing and rate-limit usage.

    Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.

    Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.

    For example, `output_tokens` will be non-zero, even for an empty string response from Claude.

    Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.

### Message Content Param

- `class MessageContentParam`

  - `string`

  - `list<ContentBlockParam>`

    - `class TextBlockParam`

      - `"text" type`

      - `string text`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?list<TextCitationParam> citations`

    - `class ImageBlockParam`

      - `"image" type`

      - `Source source`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?ImageTransformationsParam transformations`

        Configures the transformations the server applies to this image before the model observes it. Each key names a condition the server transforms images for; its value selects the transformation applied. Omitted keys keep their default behavior, and an empty object is equivalent to omitting the field.

    - `class DocumentBlockParam`

      - `"document" type`

      - `Source source`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?CitationsConfigParam citations`

      - `?string context`

      - `?string title`

    - `class SearchResultBlockParam`

      - `"search_result" type`

      - `list<TextBlockParam> content`

      - `string source`

      - `string title`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?CitationsConfigParam citations`

    - `class ThinkingBlockParam`

      - `"thinking" type`

      - `string signature`

        The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.

        Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.

      - `string thinking`

        The `thinking` text of this block as returned by the API.

    - `class RedactedThinkingBlockParam`

      - `"redacted_thinking" type`

      - `string data`

        The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.

    - `class ToolUseBlockParam`

      - `"tool_use" type`

      - `string id`

      - `array<string,mixed> input`

      - `string name`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?Caller caller`

      - `?string toolsetName`

        For a toolset member tool_use, the toolset family this member belongs to.

    - `class ToolResultBlockParam`

      - `"tool_result" type`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?Content content`

      - `?bool isError`

      - `?string toolsetName`

        For a toolset member tool_result, the toolset family of the paired tool_use.

    - `class ServerToolUseBlockParam`

      - `"server_tool_use" type`

      - `string id`

      - `array<string,mixed> input`

      - `Name name`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?Caller caller`

    - `class WebSearchToolResultBlockParam`

      - `"web_search_tool_result" type`

      - `WebSearchToolResultBlockParamContent content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?Caller caller`

    - `class WebFetchToolResultBlockParam`

      - `"web_fetch_tool_result" type`

      - `Content content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

      - `?Caller caller`

    - `class CodeExecutionToolResultBlockParam`

      - `"code_execution_tool_result" type`

      - `CodeExecutionToolResultBlockParamContent content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

    - `class BashCodeExecutionToolResultBlockParam`

      - `"bash_code_execution_tool_result" type`

      - `Content content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

    - `class TextEditorCodeExecutionToolResultBlockParam`

      - `"text_editor_code_execution_tool_result" type`

      - `Content content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

    - `class ToolSearchToolResultBlockParam`

      - `"tool_search_tool_result" type`

      - `Content content`

      - `string toolUseID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

    - `class ContainerUploadBlockParam`

      - `"container_upload" type`

      - `string fileID`

      - `?CacheControlEphemeral cacheControl`

        Create a cache control breakpoint at this content block.

### Message Count Tokens Tool

- `class MessageCountTokensTool`

  - `class Tool`

    - `?Type type`

    - `InputSchema inputSchema`

      [JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.

      This defines the shape of the `input` that your tool accepts and that the model will produce.

    - `string name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?string description`

      Description of what this tool does.

      Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.

    - `?bool eagerInputStreaming`

      Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolBash20250124`

    - `"bash_20250124" type`

    - `"bash" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20250522`

    - `"code_execution_20250522" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20250825`

    - `"code_execution_20250825" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20260120`

    - `"code_execution_20260120" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20260521`

    - `"code_execution_20260521" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class BrowserToolset20260801`

    - `"browser_toolset_20260801" type`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?BrowserToolsetConfigs configs`

      Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

  - `class MemoryTool20250818`

    - `"memory_20250818" type`

    - `"memory" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ComputerToolset20260801`

    - `"computer_toolset_20260801" type`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?ComputerToolsetConfigs configs`

      Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

  - `class ToolTextEditor20250124`

    - `"text_editor_20250124" type`

    - `"str_replace_editor" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolTextEditor20250429`

    - `"text_editor_20250429" type`

    - `"str_replace_based_edit_tool" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolTextEditor20250728`

    - `"text_editor_20250728" type`

    - `"str_replace_based_edit_tool" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?int maxCharacters`

      Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class WebSearchTool20250305`

    - `"web_search_20250305" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20250910`

    - `"web_fetch_20250910" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `class WebSearchTool20260209`

    - `"web_search_20260209" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20260209`

    - `"web_fetch_20260209" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `class WebFetchTool20260309`

    - `"web_fetch_20260309" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

    - `?bool useCache`

      Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

  - `class WebSearchTool20260318`

    - `"web_search_20260318" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?ResponseInclusion responseInclusion`

      How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20260318`

    - `"web_fetch_20260318" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?ResponseInclusion responseInclusion`

      How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

    - `?bool useCache`

      Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

  - `class ToolSearchToolBm25_20251119`

    - `Type type`

    - `"tool_search_tool_bm25" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolSearchToolRegex20251119`

    - `Type type`

    - `"tool_search_tool_regex" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

### Message Create Params Container

- `class MessageCreateParamsContainer`

  - `class ContainerParams`

    - `?string id`

      Container id

    - `?list<SkillParams> skills`

      List of skills to load in the container

  - `string`

### Message Delta Usage

- `class MessageDeltaUsage`

  - `?int cacheCreationInputTokens`

    The cumulative number of input tokens used to create the cache entry.

  - `?int cacheReadInputTokens`

    The cumulative number of input tokens read from the cache.

  - `?int inputTokens`

    The cumulative number of input tokens which were used.

  - `int outputTokens`

    The cumulative number of output tokens which were used.

  - `?OutputTokensDetails outputTokensDetails`

    Breakdown of output tokens by category.

    `output_tokens` remains the inclusive, authoritative total used for billing.
    This object provides a read-only decomposition for observability — for example,
    how many of the billed output tokens were spent on internal reasoning that may
    have been summarized before being returned to you.

  - `?ServerToolUsage serverToolUse`

    The number of server tool requests.

### Message Param

- `class MessageParam`

  - `class UserMessageParam`

    - `MessageContentParam content`

    - `"user" role`

  - `class AssistantMessageParam`

    - `MessageContentParam content`

    - `"assistant" role`

  - `class SystemMessageParam`

    - `Content content`

    - `"system" role`

### Message Tokens Count

- `class MessageTokensCount`

  - `int inputTokens`

    The total number of tokens across the provided list of messages, system prompt, and tools.

### Metadata

- `class Metadata`

  - `?string userID`

    An external identifier for the user who is associated with the request.

    This should be a uuid, hash value, or other opaque identifier. Anthropic may use this id to help detect abuse. Do not include any identifying information such as name, email address, or phone number.

### Model

- `enum Model`

  - `"claude-haiku-5-5"`

    Fastest model for high-volume, real-time tasks

  - `"claude-sonnet-5-5"`

    Efficient model for coding and agents

  - `"claude-fable-5-1"`

    Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows

  - `"claude-opus-5-5"`

    Powerful intelligence for coding, knowledge work, and long-running agents

  - `"claude-mythos-5-1"`

    Our most capable model for cybersecurity and biology research, available through trusted access programs

  - `"claude-sonnet-5"`

    Efficient model for coding and agents

  - `"claude-fable-5"`

    Next generation of intelligence for the hardest knowledge work and coding problems

  - `"claude-mythos-5"`

    Most capable model for cybersecurity and biology research

  - `"claude-opus-5"`

    Powerful intelligence for long-running agents and coding

  - `"claude-opus-4-8"`

    Powerful intelligence for long-running agents and coding

  - `"claude-opus-4-7"`

    Powerful intelligence for long-running agents and coding

  - `"claude-opus-4-6"`

    Powerful intelligence for long-running agents and coding

  - `"claude-sonnet-4-6"`

    Best combination of speed and intelligence

  - `"claude-haiku-4-5"`

  - `"claude-haiku-4-5-20251001"`

  - `"claude-opus-4-5"`

    Powerful intelligence for long-running agents and coding

  - `"claude-opus-4-5-20251101"`

    Powerful intelligence for long-running agents and coding

  - `"claude-mythos-preview"`

    **Deprecated**: Will reach end-of-life on June 30, 2026. Please migrate to claude-mythos-5. Visit https://docs.anthropic.com/en/docs/resources/model-deprecations for more information.

    New class of intelligence, strongest in coding and cybersecurity

  - `"claude-sonnet-4-5"`

    **Deprecated**: Will reach end-of-life on November 30, 2026. Please migrate to claude-sonnet-5-5. Visit https://docs.anthropic.com/en/docs/resources/model-deprecations for more information.

    High-performance model for agents and coding

  - `"claude-sonnet-4-5-20250929"`

    **Deprecated**: Will reach end-of-life on November 30, 2026. Please migrate to claude-sonnet-5-5. Visit https://docs.anthropic.com/en/docs/resources/model-deprecations for more information.

    High-performance model for agents and coding

  - `string`

### Output Config

- `class OutputConfig`

  - `?Effort effort`

    How much effort the model should put into its response. Higher effort levels may result in more thorough analysis but take longer.

    Valid values are `low`, `medium`, `high`, `xhigh`, or `max`.

  - `?JSONOutputFormat format`

    A schema to specify Claude's output format in responses. See [structured outputs](https://platform.claude.com/docs/en/build-with-claude/structured-outputs?sdk_languages=php)

### Output Tokens Details

- `class OutputTokensDetails`

  - `int thinkingTokens`

    Number of output tokens the model generated as internal reasoning, including
    the thinking-block delimiter tokens.

    Reflects the raw reasoning the model produced, not the (possibly shorter)
    summarized thinking text returned in the response body. It may differ from the model's exact
    generation count by a small number of tokens. Always ≤ `output_tokens`;
    `output_tokens - thinking_tokens` approximates the non-reasoning output.

### Plain Text Source

- `class PlainTextSource`

  - `"text" type`

  - `string data`

  - `"text/plain" mediaType`

### Raw Content Block Delta

- `class RawContentBlockDelta`

  - `class TextDelta`

    - `"text_delta" type`

    - `string text`

  - `class InputJSONDelta`

    - `"input_json_delta" type`

    - `string partialJSON`

  - `class CitationsDelta`

    - `"citations_delta" type`

    - `Citation citation`

  - `class ThinkingDelta`

    - `"thinking_delta" type`

    - `string thinking`

      The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.

  - `class SignatureDelta`

    - `"signature_delta" type`

    - `string signature`

      The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.

### Raw Content Block Delta Event

- `class RawContentBlockDeltaEvent`

  - `"content_block_delta" type`

  - `RawContentBlockDelta delta`

  - `int index`

### Raw Content Block Start Event

- `class RawContentBlockStartEvent`

  - `"content_block_start" type`

  - `ContentBlock contentBlock`

  - `int index`

### Raw Content Block Stop Event

- `class RawContentBlockStopEvent`

  - `"content_block_stop" type`

  - `int index`

### Raw Message Delta Event

- `class RawMessageDeltaEvent`

  - `"message_delta" type`

  - `Delta delta`

  - `MessageDeltaUsage usage`

    Billing and rate-limit usage.

    Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.

    Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.

    For example, `output_tokens` will be non-zero, even for an empty string response from Claude.

    Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.

### Raw Message Start Event

- `class RawMessageStartEvent`

  - `"message_start" type`

  - `Message message`

### Raw Message Stop Event

- `class RawMessageStopEvent`

  - `"message_stop" type`

### Raw Message Stream Event

- `class RawMessageStreamEvent`

  - `class RawMessageStartEvent`

    - `"message_start" type`

    - `Message message`

  - `class RawMessageDeltaEvent`

    - `"message_delta" type`

    - `Delta delta`

    - `MessageDeltaUsage usage`

      Billing and rate-limit usage.

      Anthropic's API bills and rate-limits by token counts, as tokens represent the underlying cost to our systems.

      Under the hood, the API transforms requests into a format suitable for the model. The model's output then goes through a parsing stage before becoming an API response. As a result, the token counts in `usage` will not match one-to-one with the exact visible content of an API request or response.

      For example, `output_tokens` will be non-zero, even for an empty string response from Claude.

      Total input tokens in a request is the summation of `input_tokens`, `cache_creation_input_tokens`, and `cache_read_input_tokens`.

  - `class RawMessageStopEvent`

    - `"message_stop" type`

  - `class RawContentBlockStartEvent`

    - `"content_block_start" type`

    - `ContentBlock contentBlock`

    - `int index`

  - `class RawContentBlockDeltaEvent`

    - `"content_block_delta" type`

    - `RawContentBlockDelta delta`

    - `int index`

  - `class RawContentBlockStopEvent`

    - `"content_block_stop" type`

    - `int index`

### Redacted Thinking Block

- `class RedactedThinkingBlock`

  - `"redacted_thinking" type`

  - `string data`

    The contents of this redacted thinking block, returned when portions of the model's thinking were safety-redacted. This field is opaque and encrypted, with no readable content.

    Pass `redacted_thinking` blocks back to the API unchanged when continuing a multi-turn conversation.

    See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php#redacted-thinking-blocks) for details.

### Redacted Thinking Block Param

- `class RedactedThinkingBlockParam`

  - `"redacted_thinking" type`

  - `string data`

    The `data` value of this redacted thinking block, exactly as returned by the API in a previous response. Opaque and encrypted; pass it back unchanged.

### Refusal Stop Details

- `class RefusalStopDetails`

  - `"refusal" type`

  - `?Category category`

    The policy category that triggered the refusal.

    `null` when the refusal doesn't map to a named category.

  - `?string explanation`

    Human-readable explanation of the refusal.

    This text is not guaranteed to be stable. `null` when no explanation is available for the category.

### Search Result Block Param

- `class SearchResultBlockParam`

  - `"search_result" type`

  - `list<TextBlockParam> content`

  - `string source`

  - `string title`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

### Server Tool Caller

- `class ServerToolCaller`

  - `"code_execution_20250825" type`

  - `string toolID`

### Server Tool Caller 20260120

- `class ServerToolCaller20260120`

  - `"code_execution_20260120" type`

  - `string toolID`

### Server Tool Usage

- `class ServerToolUsage`

  - `int webFetchRequests`

    The number of web fetch tool requests.

  - `int webSearchRequests`

    The number of web search tool requests.

### Server Tool Use Block

- `class ServerToolUseBlock`

  - `"server_tool_use" type`

  - `string id`

  - `Caller caller`

  - `array<string,mixed> input`

  - `Name name`

### Server Tool Use Block Param

- `class ServerToolUseBlockParam`

  - `"server_tool_use" type`

  - `string id`

  - `array<string,mixed> input`

  - `Name name`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?Caller caller`

### Signature Delta

- `class SignatureDelta`

  - `"signature_delta" type`

  - `string signature`

    The `signature` for this thinking block: an opaque value used to verify that the block was generated by Claude when it is passed back to the API. Delivered in a `signature_delta` event just before the block's `content_block_stop` event.

### Skill Params

- `class SkillParams`

  - `Type type`

    Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)

  - `string skillID`

    Skill ID

  - `?string version`

    Skill version or 'latest' for most recent version

### Stop Reason

- `enum StopReason`

  - `"end_turn"`

  - `"max_tokens"`

  - `"stop_sequence"`

  - `"tool_use"`

  - `"pause_turn"`

  - `"refusal"`

  - `"model_context_window_exceeded"`

### System Message Param

- `class SystemMessageParam`

  - `Content content`

  - `"system" role`

### Text Block

- `class TextBlock`

  - `"text" type`

  - `?list<TextCitation> citations`

    Citations supporting the text block.

    The type of citation returned will depend on the type of document being cited. Citing a PDF results in `page_location`, plain text results in `char_location`, and content document results in `content_block_location`.

  - `string text`

### Text Block Param

- `class TextBlockParam`

  - `"text" type`

  - `string text`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?list<TextCitationParam> citations`

### Text Citation

- `class TextCitation`

  - `class CitationCharLocation`

    - `"char_location" type`

    - `string citedText`

    - `int documentIndex`

    - `?string documentTitle`

    - `int endCharIndex`

    - `?string fileID`

    - `int startCharIndex`

  - `class CitationPageLocation`

    - `"page_location" type`

    - `string citedText`

    - `int documentIndex`

    - `?string documentTitle`

    - `int endPageNumber`

    - `?string fileID`

    - `int startPageNumber`

  - `class CitationContentBlockLocation`

    - `"content_block_location" type`

    - `string citedText`

      The full text of the cited block range, concatenated.

      Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

    - `int documentIndex`

    - `?string documentTitle`

    - `int endBlockIndex`

      Exclusive 0-based end index of the cited block range in the source's `content` array.

      Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

    - `?string fileID`

    - `int startBlockIndex`

      0-based index of the first cited block in the source's `content` array.

  - `class CitationsWebSearchResultLocation`

    - `"web_search_result_location" type`

    - `string citedText`

    - `string encryptedIndex`

    - `?string title`

    - `string url`

  - `class CitationsSearchResultLocation`

    - `"search_result_location" type`

    - `string citedText`

      The full text of the cited block range, concatenated.

      Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

    - `int endBlockIndex`

      Exclusive 0-based end index of the cited block range in the source's `content` array.

      Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

    - `int searchResultIndex`

      0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.

      Counted separately from `document_index`; server-side web search results are not included in this count.

    - `string source`

    - `int startBlockIndex`

      0-based index of the first cited block in the source's `content` array.

    - `?string title`

### Text Citation Param

- `class TextCitationParam`

  - `class CitationCharLocationParam`

    - `"char_location" type`

    - `string citedText`

    - `int documentIndex`

    - `?string documentTitle`

    - `int endCharIndex`

    - `int startCharIndex`

  - `class CitationPageLocationParam`

    - `"page_location" type`

    - `string citedText`

    - `int documentIndex`

    - `?string documentTitle`

    - `int endPageNumber`

    - `int startPageNumber`

  - `class CitationContentBlockLocationParam`

    - `"content_block_location" type`

    - `string citedText`

      The full text of the cited block range, concatenated.

      Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

    - `int documentIndex`

    - `?string documentTitle`

    - `int endBlockIndex`

      Exclusive 0-based end index of the cited block range in the source's `content` array.

      Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

    - `int startBlockIndex`

      0-based index of the first cited block in the source's `content` array.

  - `class CitationWebSearchResultLocationParam`

    - `"web_search_result_location" type`

    - `string citedText`

    - `string encryptedIndex`

    - `?string title`

    - `string url`

  - `class CitationSearchResultLocationParam`

    - `"search_result_location" type`

    - `string citedText`

      The full text of the cited block range, concatenated.

      Always equals the contents of `content[start_block_index:end_block_index]` joined together. The text block is the minimal citable unit; this field is never a substring of a single block. Not counted toward output tokens, and not counted toward input tokens when sent back in subsequent turns.

    - `int endBlockIndex`

      Exclusive 0-based end index of the cited block range in the source's `content` array.

      Always greater than `start_block_index`; a single-block citation has `end_block_index = start_block_index + 1`.

    - `int searchResultIndex`

      0-based index of the cited search result among all `search_result` content blocks in the request, in the order they appear across messages and tool results.

      Counted separately from `document_index`; server-side web search results are not included in this count.

    - `string source`

    - `int startBlockIndex`

      0-based index of the first cited block in the source's `content` array.

    - `?string title`

### Text Delta

- `class TextDelta`

  - `"text_delta" type`

  - `string text`

### Text Editor Code Execution Create Result Block

- `class TextEditorCodeExecutionCreateResultBlock`

  - `"text_editor_code_execution_create_result" type`

  - `bool isFileUpdate`

### Text Editor Code Execution Create Result Block Param

- `class TextEditorCodeExecutionCreateResultBlockParam`

  - `"text_editor_code_execution_create_result" type`

  - `bool isFileUpdate`

### Text Editor Code Execution Str Replace Result Block

- `class TextEditorCodeExecutionStrReplaceResultBlock`

  - `"text_editor_code_execution_str_replace_result" type`

  - `?list<string> lines`

  - `?int newLines`

  - `?int newStart`

  - `?int oldLines`

  - `?int oldStart`

### Text Editor Code Execution Str Replace Result Block Param

- `class TextEditorCodeExecutionStrReplaceResultBlockParam`

  - `"text_editor_code_execution_str_replace_result" type`

  - `?list<string> lines`

  - `?int newLines`

  - `?int newStart`

  - `?int oldLines`

  - `?int oldStart`

### Text Editor Code Execution Tool Result Block

- `class TextEditorCodeExecutionToolResultBlock`

  - `"text_editor_code_execution_tool_result" type`

  - `Content content`

  - `string toolUseID`

### Text Editor Code Execution Tool Result Block Param

- `class TextEditorCodeExecutionToolResultBlockParam`

  - `"text_editor_code_execution_tool_result" type`

  - `Content content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Text Editor Code Execution Tool Result Error

- `class TextEditorCodeExecutionToolResultError`

  - `"text_editor_code_execution_tool_result_error" type`

  - `TextEditorCodeExecutionToolResultErrorCode errorCode`

  - `?string errorMessage`

### Text Editor Code Execution Tool Result Error Code

- `enum TextEditorCodeExecutionToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"unavailable"`

  - `"too_many_requests"`

  - `"execution_time_exceeded"`

  - `"file_not_found"`

### Text Editor Code Execution Tool Result Error Param

- `class TextEditorCodeExecutionToolResultErrorParam`

  - `"text_editor_code_execution_tool_result_error" type`

  - `TextEditorCodeExecutionToolResultErrorCode errorCode`

  - `?string errorMessage`

### Text Editor Code Execution View Result Block

- `class TextEditorCodeExecutionViewResultBlock`

  - `"text_editor_code_execution_view_result" type`

  - `string content`

  - `FileType fileType`

  - `?int numLines`

  - `?int startLine`

  - `?int totalLines`

### Text Editor Code Execution View Result Block Param

- `class TextEditorCodeExecutionViewResultBlockParam`

  - `"text_editor_code_execution_view_result" type`

  - `string content`

  - `FileType fileType`

  - `?int numLines`

  - `?int startLine`

  - `?int totalLines`

### Thinking Block

- `class ThinkingBlock`

  - `"thinking" type`

  - `string signature`

    A value used to verify that this thinking block was generated by Claude when it is passed back to the API.

    This is an opaque field and should not be interpreted or parsed. When passing thinking blocks back to the API (required when using tools with extended thinking), pass them back exactly as received, with this field intact.

    See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php) for details.

  - `string thinking`

    The text of Claude's thinking process for this block.

### Thinking Block Param

- `class ThinkingBlockParam`

  - `"thinking" type`

  - `string signature`

    The `signature` value of this thinking block, exactly as returned by the API in a previous response. Used to verify that the block was generated by Claude.

    Thinking blocks must be passed back unmodified and in their original order; a modified block results in a 400 `invalid_request_error`.

  - `string thinking`

    The `thinking` text of this block as returned by the API.

### Thinking Config Adaptive

- `class ThinkingConfigAdaptive`

  - `"adaptive" type`

  - `?Display display`

    Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. The default depends on the model; see [Controlling thinking display](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#controlling-thinking-display).

### Thinking Config Between Tools

- `class ThinkingConfigBetweenTools`

  - `"between_tools" type`

### Thinking Config Disabled

- `class ThinkingConfigDisabled`

  - `"disabled" type`

### Thinking Config Enabled

- `class ThinkingConfigEnabled`

  - `"enabled" type`

  - `int budgetTokens`

    Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.

    Must be ≥1024 and less than `max_tokens`.

    See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php) for details.

  - `?Display display`

    Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. The default depends on the model; see [Controlling thinking display](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#controlling-thinking-display).

### Thinking Config Param

- `class ThinkingConfigParam`

  - `class ThinkingConfigEnabled`

    - `"enabled" type`

    - `int budgetTokens`

      Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality.

      Must be ≥1024 and less than `max_tokens`.

      See [extended thinking](https://platform.claude.com/docs/en/build-with-claude/extended-thinking?sdk_languages=php) for details.

    - `?Display display`

      Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. The default depends on the model; see [Controlling thinking display](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#controlling-thinking-display).

  - `class ThinkingConfigDisabled`

    - `"disabled" type`

  - `class ThinkingConfigBetweenTools`

    - `"between_tools" type`

  - `class ThinkingConfigAdaptive`

    - `"adaptive" type`

    - `?Display display`

      Controls how thinking content appears in the response. When set to `summarized`, thinking is returned normally. When set to `omitted`, thinking content is redacted but a signature is returned for multi-turn continuity. The default depends on the model; see [Controlling thinking display](https://platform.claude.com/docs/en/build-with-claude/thinking?sdk_languages=php#controlling-thinking-display).

### Thinking Delta

- `class ThinkingDelta`

  - `"thinking_delta" type`

  - `string thinking`

    The incremental `thinking` text for this content block. Concatenate the `thinking` values of successive `thinking_delta` events to assemble the block's full `thinking` value.

### Tool

- `class Tool`

  - `?Type type`

  - `InputSchema inputSchema`

    [JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.

    This defines the shape of the `input` that your tool accepts and that the model will produce.

  - `string name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?string description`

    Description of what this tool does.

    Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.

  - `?bool eagerInputStreaming`

    Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.

  - `?list<array<string,mixed>> inputExamples`

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Bash 20250124

- `class ToolBash20250124`

  - `"bash_20250124" type`

  - `"bash" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?list<array<string,mixed>> inputExamples`

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Choice

- `class ToolChoice`

  - `class ToolChoiceAuto`

    - `"auto" type`

    - `?bool disableParallelToolUse`

      Whether to disable parallel tool use.

      Defaults to `false`. If set to `true`, the model will output at most one tool use.

  - `class ToolChoiceAny`

    - `"any" type`

    - `?bool disableParallelToolUse`

      Whether to disable parallel tool use.

      Defaults to `false`. If set to `true`, the model will output exactly one tool use.

  - `class ToolChoiceTool`

    - `"tool" type`

    - `string name`

      The name of the tool to use.

    - `?bool disableParallelToolUse`

      Whether to disable parallel tool use.

      Defaults to `false`. If set to `true`, the model will output exactly one tool use.

  - `class ToolChoiceNone`

    - `"none" type`

### Tool Choice Any

- `class ToolChoiceAny`

  - `"any" type`

  - `?bool disableParallelToolUse`

    Whether to disable parallel tool use.

    Defaults to `false`. If set to `true`, the model will output exactly one tool use.

### Tool Choice Auto

- `class ToolChoiceAuto`

  - `"auto" type`

  - `?bool disableParallelToolUse`

    Whether to disable parallel tool use.

    Defaults to `false`. If set to `true`, the model will output at most one tool use.

### Tool Choice None

- `class ToolChoiceNone`

  - `"none" type`

### Tool Choice Tool

- `class ToolChoiceTool`

  - `"tool" type`

  - `string name`

    The name of the tool to use.

  - `?bool disableParallelToolUse`

    Whether to disable parallel tool use.

    Defaults to `false`. If set to `true`, the model will output exactly one tool use.

### Tool Reference Block

- `class ToolReferenceBlock`

  - `"tool_reference" type`

  - `string toolName`

### Tool Reference Block Param

- `class ToolReferenceBlockParam`

  - `"tool_reference" type`

  - `string toolName`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Tool Result Block Param

- `class ToolResultBlockParam`

  - `"tool_result" type`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?Content content`

  - `?bool isError`

  - `?string toolsetName`

    For a toolset member tool_result, the toolset family of the paired tool_use.

### Tool Search Tool Bm25 20251119

- `class ToolSearchToolBm25_20251119`

  - `Type type`

  - `"tool_search_tool_bm25" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Search Tool Regex 20251119

- `class ToolSearchToolRegex20251119`

  - `Type type`

  - `"tool_search_tool_regex" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Search Tool Result Block

- `class ToolSearchToolResultBlock`

  - `"tool_search_tool_result" type`

  - `Content content`

  - `string toolUseID`

### Tool Search Tool Result Block Param

- `class ToolSearchToolResultBlockParam`

  - `"tool_search_tool_result" type`

  - `Content content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

### Tool Search Tool Result Error

- `class ToolSearchToolResultError`

  - `"tool_search_tool_result_error" type`

  - `ToolSearchToolResultErrorCode errorCode`

  - `?string errorMessage`

### Tool Search Tool Result Error Code

- `enum ToolSearchToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"unavailable"`

  - `"too_many_requests"`

  - `"execution_time_exceeded"`

### Tool Search Tool Result Error Param

- `class ToolSearchToolResultErrorParam`

  - `"tool_search_tool_result_error" type`

  - `ToolSearchToolResultErrorCode errorCode`

  - `?string errorMessage`

### Tool Search Tool Search Result Block

- `class ToolSearchToolSearchResultBlock`

  - `"tool_search_tool_search_result" type`

  - `list<ToolReferenceBlock> toolReferences`

### Tool Search Tool Search Result Block Param

- `class ToolSearchToolSearchResultBlockParam`

  - `"tool_search_tool_search_result" type`

  - `list<ToolReferenceBlockParam> toolReferences`

### Tool Text Editor 20250124

- `class ToolTextEditor20250124`

  - `"text_editor_20250124" type`

  - `"str_replace_editor" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?list<array<string,mixed>> inputExamples`

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Text Editor 20250429

- `class ToolTextEditor20250429`

  - `"text_editor_20250429" type`

  - `"str_replace_based_edit_tool" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?list<array<string,mixed>> inputExamples`

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Text Editor 20250728

- `class ToolTextEditor20250728`

  - `"text_editor_20250728" type`

  - `"str_replace_based_edit_tool" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?list<array<string,mixed>> inputExamples`

  - `?int maxCharacters`

    Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

### Tool Union

- `class ToolUnion`

  - `class Tool`

    - `?Type type`

    - `InputSchema inputSchema`

      [JSON schema](https://json-schema.org/draft/2020-12) for this tool's input.

      This defines the shape of the `input` that your tool accepts and that the model will produce.

    - `string name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?string description`

      Description of what this tool does.

      Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.

    - `?bool eagerInputStreaming`

      Enable eager input streaming for this tool. When true, tool input parameters will be streamed incrementally as they are generated, and types will be inferred on-the-fly rather than buffering the full JSON output. When false, streaming is disabled for this tool even if the fine-grained-tool-streaming beta is active. When null (default), uses the default behavior based on beta headers.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolBash20250124`

    - `"bash_20250124" type`

    - `"bash" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20250522`

    - `"code_execution_20250522" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20250825`

    - `"code_execution_20250825" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20260120`

    - `"code_execution_20260120" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class CodeExecutionTool20260521`

    - `"code_execution_20260521" type`

    - `"code_execution" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class BrowserToolset20260801`

    - `"browser_toolset_20260801" type`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?BrowserToolsetConfigs configs`

      Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

  - `class MemoryTool20250818`

    - `"memory_20250818" type`

    - `"memory" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ComputerToolset20260801`

    - `"computer_toolset_20260801" type`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?ComputerToolsetConfigs configs`

      Sparse per-member overrides, keyed by member name. Absent, null, and {} are equivalent; a member's defaults apply wherever its key is absent.

  - `class ToolTextEditor20250124`

    - `"text_editor_20250124" type`

    - `"str_replace_editor" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolTextEditor20250429`

    - `"text_editor_20250429" type`

    - `"str_replace_based_edit_tool" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolTextEditor20250728`

    - `"text_editor_20250728" type`

    - `"str_replace_based_edit_tool" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?list<array<string,mixed>> inputExamples`

    - `?int maxCharacters`

      Maximum number of characters to display when viewing a file. If not specified, defaults to displaying the full file.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class WebSearchTool20250305`

    - `"web_search_20250305" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20250910`

    - `"web_fetch_20250910" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `class WebSearchTool20260209`

    - `"web_search_20260209" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20260209`

    - `"web_fetch_20260209" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `class WebFetchTool20260309`

    - `"web_fetch_20260309" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

    - `?bool useCache`

      Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

  - `class WebSearchTool20260318`

    - `"web_search_20260318" type`

    - `"web_search" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

    - `?list<string> blockedDomains`

      If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?ResponseInclusion responseInclusion`

      How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?UserLocation userLocation`

      Parameters for the user's location. Used to provide more relevant search results.

  - `class WebFetchTool20260318`

    - `"web_fetch_20260318" type`

    - `"web_fetch" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?list<string> allowedDomains`

      List of domains to allow fetching from

    - `?list<string> blockedDomains`

      List of domains to block fetching from

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?CitationsConfigParam citations`

      Citations configuration for fetched documents. Citations are disabled by default.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?int maxContentTokens`

      Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

    - `?int maxUses`

      Maximum number of times the tool can be used in the API request.

    - `?ResponseInclusion responseInclusion`

      How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

    - `?WebFetchURLSources urlSources`

      Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

    - `?bool useCache`

      Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

  - `class ToolSearchToolBm25_20251119`

    - `Type type`

    - `"tool_search_tool_bm25" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

  - `class ToolSearchToolRegex20251119`

    - `Type type`

    - `"tool_search_tool_regex" name`

      Name of the tool.

      This is how the tool will be called by the model and in `tool_use` blocks.

    - `?list<AllowedCaller> allowedCallers`

    - `?CacheControlEphemeral cacheControl`

      Create a cache control breakpoint at this content block.

    - `?bool deferLoading`

      If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

    - `?bool strict`

      When true, guarantees schema validation on tool names and inputs

### Tool Use Block

- `class ToolUseBlock`

  - `"tool_use" type`

  - `string id`

  - `Caller caller`

  - `array<string,mixed> input`

  - `string name`

  - `?string toolsetName`

    For a toolset member tool_use, the toolset family.

### Tool Use Block Param

- `class ToolUseBlockParam`

  - `"tool_use" type`

  - `string id`

  - `array<string,mixed> input`

  - `string name`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?Caller caller`

  - `?string toolsetName`

    For a toolset member tool_use, the toolset family this member belongs to.

### Tool Use Caller

- `class ToolUseCaller`

  - `class DirectCaller`

    - `"direct" type`

  - `class ServerToolCaller`

    - `"code_execution_20250825" type`

    - `string toolID`

  - `class ServerToolCaller20260120`

    - `"code_execution_20260120" type`

    - `string toolID`

### Toolset Tool Use Block

- `class ToolsetToolUseBlock`

  - `class BrowserToolUseBlock`

    - `class BrowserNavigateToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserNavigateInput input`

        Navigate to a URL, or go back/forward/reload in history. The protocol may be
        omitted (defaults to https://).

      - `"navigate" name`

      - `"browser" toolsetName`

    - `class BrowserListTabsToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserListTabsInput input`

        List all open tabs with each tab's tab_id, title, and URL.

      - `"list_tabs" name`

      - `"browser" toolsetName`

    - `class BrowserNewTabToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserNewTabInput input`

        Open a new empty tab and return its tab_id.

      - `"new_tab" name`

      - `"browser" toolsetName`

    - `class BrowserSwitchTabToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserSwitchTabInput input`

        Make the tab with the given tab_id the active tab — the tab that actions without
        a tab_id apply to.

      - `"switch_tab" name`

      - `"browser" toolsetName`

    - `class BrowserCloseTabToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserCloseTabInput input`

        Close the tab with the given tab_id.

      - `"close_tab" name`

      - `"browser" toolsetName`

    - `class BrowserReadPageToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserReadPageInput input`

        Return a structured accessibility tree of the page (or the subtree rooted at
        `ref`), with element references like [ref_7] that can be used as targets on later
        actions. Output is capped at 50,000 characters — narrow with `ref` or a smaller
        `depth` when exceeded.

      - `"read_page" name`

      - `"browser" toolsetName`

    - `class BrowserGetPageTextToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserGetPageTextInput input`

        Return the page's visible text content as plain text, prioritizing article
        content. Suited to articles, documentation, and other text-heavy pages.

      - `"get_page_text" name`

      - `"browser" toolsetName`

    - `class BrowserReadConsoleToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserReadConsoleInput input`

        Return console output (log entries, errors, warnings) accumulated since the
        driver attached to the tab and since the last read, one line per entry. An empty
        result does not mean no traffic for a tab that predates attach.

      - `"read_console" name`

      - `"browser" toolsetName`

    - `class BrowserReadNetworkToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserReadNetworkInput input`

        Return the network requests (method, URL, status, MIME type, timing) recorded
        since the driver attached to the tab and since the last read, one line per entry.
        An empty result does not mean no traffic for a tab that predates attach.

      - `"read_network" name`

      - `"browser" toolsetName`

    - `class BrowserFindToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserFindInput input`

        Find elements matching a natural-language description (e.g. "search bar", "add to
        cart button") and return up to 20 matches with element references.

      - `"find" name`

      - `"browser" toolsetName`

    - `class BrowserFormInputToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserFormInputInput input`

        Set the value of a form element (input, textarea, select, checkbox). Use a
        boolean for checkboxes, an option value or text for selects.

      - `"form_input" name`

      - `"browser" toolsetName`

    - `class BrowserFileUploadToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserFileUploadInput input`

        Set the value of a file-input element to one or more files. The target must be an
        element reference; at least one of paths or document_ids is required.

      - `"file_upload" name`

      - `"browser" toolsetName`

    - `class BrowserScrollToToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserScrollToInput input`

        Scroll an element into view.

      - `"scroll_to" name`

      - `"browser" toolsetName`

    - `class BrowserScreenshotToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserScreenshotInput input`

        Capture the current browser viewport.

      - `"screenshot" name`

      - `"browser" toolsetName`

    - `class BrowserZoomToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserZoomInput input`

        Return a cropped screenshot of the given viewport region, scaled up for closer
        inspection — useful for small icons, buttons, or text. Coordinates are in the
        same viewport-pixel space as a full screenshot.

      - `"zoom" name`

      - `"browser" toolsetName`

    - `class BrowserLeftClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserLeftClickInput input`

        Left-click at a viewport coordinate or on an element by reference.

      - `"left_click" name`

      - `"browser" toolsetName`

    - `class BrowserRightClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserRightClickInput input`

        Right-click at a viewport coordinate or on an element by reference.

      - `"right_click" name`

      - `"browser" toolsetName`

    - `class BrowserMiddleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserMiddleClickInput input`

        Middle-click at a viewport coordinate or on an element by reference.

      - `"middle_click" name`

      - `"browser" toolsetName`

    - `class BrowserDoubleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserDoubleClickInput input`

        Double left-click at a viewport coordinate or on an element by reference.

      - `"double_click" name`

      - `"browser" toolsetName`

    - `class BrowserTripleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserTripleClickInput input`

        Triple left-click at a viewport coordinate or on an element by reference
        (typically selects a line or paragraph).

      - `"triple_click" name`

      - `"browser" toolsetName`

    - `class BrowserHoverToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserHoverInput input`

        Move the cursor to a coordinate or element without clicking.

      - `"hover" name`

      - `"browser" toolsetName`

    - `class BrowserLeftClickDragToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserLeftClickDragInput input`

        Press at `from`, drag to `target`, release. Both must be coordinate targets.

      - `"left_click_drag" name`

      - `"browser" toolsetName`

    - `class BrowserLeftMouseDownToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserLeftMouseDownInput input`

        Press and hold the left mouse button at a viewport coordinate. Pair with
        left_mouse_up to perform a custom drag.

      - `"left_mouse_down" name`

      - `"browser" toolsetName`

    - `class BrowserLeftMouseUpToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserLeftMouseUpInput input`

        Release the left mouse button at a viewport coordinate.

      - `"left_mouse_up" name`

      - `"browser" toolsetName`

    - `class BrowserMouseMoveToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserMouseMoveInput input`

        Move the pointer to a viewport coordinate without clicking.

      - `"mouse_move" name`

      - `"browser" toolsetName`

    - `class BrowserScrollToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserScrollInput input`

        Scroll at a viewport position. `target` must be a coordinate target.

      - `"scroll" name`

      - `"browser" toolsetName`

    - `class BrowserTypeToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserTypeInput input`

        Type a literal string at the current focus.

      - `"type" name`

      - `"browser" toolsetName`

    - `class BrowserKeyToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserKeyInput input`

        Press a key or key chord. Use "+" to combine modifiers with a key (e.g. "ctrl+a",
        "cmd+shift+p") and space to sequence presses (e.g. "Backspace Backspace Delete").
        Common names like "Return", "Tab", "Escape", "BackSpace" are supported.

      - `"key" name`

      - `"browser" toolsetName`

    - `class BrowserHoldKeyToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserHoldKeyInput input`

        Hold a key or key chord down for a duration, then release it. Uses the same key
        names and "+" chord syntax as the key action.

      - `"hold_key" name`

      - `"browser" toolsetName`

    - `class BrowserWaitToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserWaitInput input`

        Pause for the given duration.

      - `"wait" name`

      - `"browser" toolsetName`

    - `class BrowserJavascriptExecToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `BrowserJavascriptExecInput input`

        Execute JavaScript in the page context and return the value of the last
        expression. The code runs with access to the DOM, `window`, and page variables.
        Write the expression you want evaluated — do NOT use `return`.

      - `"javascript_exec" name`

      - `"browser" toolsetName`

  - `class ComputerToolUseBlock`

    - `class ComputerKeyToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerKeyInput input`

        Press a key or key-combination on the keyboard. Use "+" to combine modifiers with
        a key (e.g. "ctrl+s", "alt+Tab", "ctrl+shift+Escape"). Key names are
        case-insensitive; common names like "Return", "Tab", "Escape", "Up", "Down",
        "Left", "Right", "Home", "End", "Page_Up", "Page_Down", "Delete", "BackSpace" are
        supported.

      - `"key" name`

      - `"computer" toolsetName`

    - `class ComputerHoldKeyToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerHoldKeyInput input`

        Hold down a key or key-combination for a specified duration. Uses the same key
        syntax as `key`.

      - `"hold_key" name`

      - `"computer" toolsetName`

    - `class ComputerTypeToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerTypeInput input`

        Type a string of text on the keyboard.

      - `"type" name`

      - `"computer" toolsetName`

    - `class ComputerCursorPositionToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerCursorPositionInput input`

        Get the current (x, y) pixel coordinate of the cursor.

      - `"cursor_position" name`

      - `"computer" toolsetName`

    - `class ComputerMouseMoveToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerMouseMoveInput input`

        Move the cursor to a specified (x, y) pixel coordinate. Use this ONLY to hover
        without clicking; otherwise use a click action directly.

      - `"mouse_move" name`

      - `"computer" toolsetName`

    - `class ComputerLeftMouseDownToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerLeftMouseDownInput input`

        Press and hold the left mouse button at the current cursor position.

      - `"left_mouse_down" name`

      - `"computer" toolsetName`

    - `class ComputerLeftMouseUpToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerLeftMouseUpInput input`

        Release the left mouse button.

      - `"left_mouse_up" name`

      - `"computer" toolsetName`

    - `class ComputerLeftClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerLeftClickInput input`

        Click the left mouse button at the specified (x, y) pixel coordinate, or the
        current cursor position if `coordinate` is omitted.

      - `"left_click" name`

      - `"computer" toolsetName`

    - `class ComputerLeftClickDragToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerLeftClickDragInput input`

        Click and drag the cursor from `start_coordinate` to `coordinate`.

      - `"left_click_drag" name`

      - `"computer" toolsetName`

    - `class ComputerRightClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerRightClickInput input`

        Click the right mouse button at the specified (x, y) pixel coordinate, or the
        current cursor position if `coordinate` is omitted.

      - `"right_click" name`

      - `"computer" toolsetName`

    - `class ComputerMiddleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerMiddleClickInput input`

        Click the middle mouse button at the specified (x, y) pixel coordinate, or the
        current cursor position if `coordinate` is omitted.

      - `"middle_click" name`

      - `"computer" toolsetName`

    - `class ComputerDoubleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerDoubleClickInput input`

        Double-click the left mouse button at the specified (x, y) pixel coordinate, or
        the current cursor position if `coordinate` is omitted.

      - `"double_click" name`

      - `"computer" toolsetName`

    - `class ComputerTripleClickToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerTripleClickInput input`

        Triple-click the left mouse button at the specified (x, y) pixel coordinate, or
        the current cursor position if `coordinate` is omitted.

      - `"triple_click" name`

      - `"computer" toolsetName`

    - `class ComputerScrollToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerScrollInput input`

        Scroll the screen at the specified (x, y) pixel coordinate, or the current cursor
        position if `coordinate` is omitted. Do NOT use PageUp/PageDown to scroll.

      - `"scroll" name`

      - `"computer" toolsetName`

    - `class ComputerWaitToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerWaitInput input`

        Wait for a specified duration.

      - `"wait" name`

      - `"computer" toolsetName`

    - `class ComputerScreenshotToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerScreenshotInput input`

        Take a screenshot of the screen.

      - `"screenshot" name`

      - `"computer" toolsetName`

    - `class ComputerZoomToolUseBlock`

      - `"tool_use" type`

      - `string id`

      - `ToolUseCaller caller`

        Which party invoked the tool call: the model directly, or a server tool on its behalf.

      - `ComputerZoomInput input`

        Take a screenshot of a rectangular region. Region coordinates are in the
        full-screenshot space (not physical display pixels). The crop is scaled up to
        fill the image budget so fine details become legible.

      - `"zoom" name`

      - `"computer" toolsetName`

  - `class ToolUseBlock`

    - `"tool_use" type`

    - `string id`

    - `Caller caller`

    - `array<string,mixed> input`

    - `string name`

    - `?string toolsetName`

      For a toolset member tool_use, the toolset family.

### URL Image Source

- `class URLImageSource`

  - `"url" type`

  - `string url`

### URL PDF Source

- `class URLPDFSource`

  - `"url" type`

  - `string url`

### Usage

- `class Usage`

  - `?CacheCreation cacheCreation`

    Breakdown of cached tokens by TTL

  - `?int cacheCreationInputTokens`

    The number of input tokens used to create the cache entry.

  - `?int cacheReadInputTokens`

    The number of input tokens read from the cache.

  - `?string inferenceGeo`

    The geographic region where inference was performed for this request.

  - `int inputTokens`

    The number of input tokens which were used.

  - `int outputTokens`

    The number of output tokens which were used.

  - `?OutputTokensDetails outputTokensDetails`

    Breakdown of output tokens by category.

    `output_tokens` remains the inclusive, authoritative total used for billing.
    This object provides a read-only decomposition for observability — for example,
    how many of the billed output tokens were spent on internal reasoning that may
    have been summarized before being returned to you.

  - `?ServerToolUsage serverToolUse`

    The number of server tool requests.

  - `?ServiceTier serviceTier`

    If the request used the priority, standard, or batch tier.

  - `?Speed speed`

    The inference speed mode used for this request.

### User Location

- `class UserLocation`

  - `"approximate" type`

  - `?string city`

    The city of the user.

  - `?string country`

    The two letter [ISO country code](https://en.wikipedia.org/wiki/ISO_3166-1_alpha-2) of the user.

  - `?string region`

    The region of the user.

  - `?string timezone`

    The [IANA timezone](https://nodatime.org/TimeZones) of the user.

### User Message Param

- `class UserMessageParam`

  - `MessageContentParam content`

  - `"user" role`

### Web Fetch Block

- `class WebFetchBlock`

  - `"web_fetch_result" type`

  - `DocumentBlock content`

  - `?string retrievedAt`

    ISO 8601 timestamp when the content was retrieved

  - `string url`

    Fetched content URL

### Web Fetch Block Param

- `class WebFetchBlockParam`

  - `"web_fetch_result" type`

  - `DocumentBlockParam content`

  - `string url`

    Fetched content URL

  - `?string retrievedAt`

    ISO 8601 timestamp when the content was retrieved

### Web Fetch Tool 20250910

- `class WebFetchTool20250910`

  - `"web_fetch_20250910" type`

  - `"web_fetch" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    List of domains to allow fetching from

  - `?list<string> blockedDomains`

    List of domains to block fetching from

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

    Citations configuration for fetched documents. Citations are disabled by default.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxContentTokens`

    Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?WebFetchURLSources urlSources`

    Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

### Web Fetch Tool 20260209

- `class WebFetchTool20260209`

  - `"web_fetch_20260209" type`

  - `"web_fetch" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    List of domains to allow fetching from

  - `?list<string> blockedDomains`

    List of domains to block fetching from

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

    Citations configuration for fetched documents. Citations are disabled by default.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxContentTokens`

    Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?WebFetchURLSources urlSources`

    Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

### Web Fetch Tool 20260309

- `class WebFetchTool20260309`

  - `"web_fetch_20260309" type`

  - `"web_fetch" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    List of domains to allow fetching from

  - `?list<string> blockedDomains`

    List of domains to block fetching from

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

    Citations configuration for fetched documents. Citations are disabled by default.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxContentTokens`

    Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?WebFetchURLSources urlSources`

    Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `?bool useCache`

    Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

### Web Fetch Tool 20260318

- `class WebFetchTool20260318`

  - `"web_fetch_20260318" type`

  - `"web_fetch" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    List of domains to allow fetching from

  - `?list<string> blockedDomains`

    List of domains to block fetching from

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?CitationsConfigParam citations`

    Citations configuration for fetched documents. Citations are disabled by default.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxContentTokens`

    Maximum number of tokens used by including web page text content in the context. The limit is approximate and does not apply to binary content such as PDFs.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?ResponseInclusion responseInclusion`

    How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?WebFetchURLSources urlSources`

    Which sources contribute to the set of URLs the tool may fetch. Omitted means every source.

  - `?bool useCache`

    Whether to use cached content. Set to false to bypass the cache and fetch fresh content. Only set to false when the user explicitly requests fresh content or when fetching rapidly-changing sources.

### Web Fetch Tool Result Block

- `class WebFetchToolResultBlock`

  - `"web_fetch_tool_result" type`

  - `Caller caller`

  - `Content content`

  - `string toolUseID`

### Web Fetch Tool Result Block Param

- `class WebFetchToolResultBlockParam`

  - `"web_fetch_tool_result" type`

  - `Content content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?Caller caller`

### Web Fetch Tool Result Error Block

- `class WebFetchToolResultErrorBlock`

  - `"web_fetch_tool_result_error" type`

  - `WebFetchToolResultErrorCode errorCode`

### Web Fetch Tool Result Error Block Param

- `class WebFetchToolResultErrorBlockParam`

  - `"web_fetch_tool_result_error" type`

  - `WebFetchToolResultErrorCode errorCode`

### Web Fetch Tool Result Error Code

- `enum WebFetchToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"url_too_long"`

  - `"url_not_allowed"`

  - `"url_not_in_prior_context"`

  - `"url_not_accessible"`

  - `"unsupported_content_type"`

  - `"too_many_requests"`

  - `"max_uses_exceeded"`

  - `"unavailable"`

  - `"content_too_large"`

### Web Fetch URL Source All

- `class WebFetchURLSourceAll`

  - `"all" type`

### Web Fetch URL Source Except

- `class WebFetchURLSourceExcept`

  - `"except" type`

  - `list<WebFetchURLSourceToolReference> tools`

### Web Fetch URL Source None

- `class WebFetchURLSourceNone`

  - `"none" type`

### Web Fetch URL Source Only

- `class WebFetchURLSourceOnly`

  - `"only" type`

  - `list<WebFetchURLSourceToolReference> tools`

### Web Fetch URL Source Tool Reference

- `class WebFetchURLSourceToolReference`

  - `"tool_reference" type`

  - `string name`

### Web Fetch URL Sources

- `class WebFetchURLSources`

  - `?ClientToolResults clientToolResults`

    Which client tools' results contribute fetchable URLs: "all", "none", or an only or except list of client tool names from tools[].

  - `?ServerToolResults serverToolResults`

    Which server tools' results contribute fetchable URLs: "all", "none", or an only or except list of server tool names from tools[]; only web_search and web_fetch results ever contribute.

  - `?UserInput userInput`

    Whether URLs in user messages are fetchable: "all" or "none".

### Web Search Result Block

- `class WebSearchResultBlock`

  - `"web_search_result" type`

  - `string encryptedContent`

  - `?string pageAge`

  - `string title`

  - `string url`

### Web Search Result Block Param

- `class WebSearchResultBlockParam`

  - `"web_search_result" type`

  - `string encryptedContent`

  - `string title`

  - `string url`

  - `?string pageAge`

### Web Search Tool 20250305

- `class WebSearchTool20250305`

  - `"web_search_20250305" type`

  - `"web_search" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

  - `?list<string> blockedDomains`

    If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?UserLocation userLocation`

    Parameters for the user's location. Used to provide more relevant search results.

### Web Search Tool 20260209

- `class WebSearchTool20260209`

  - `"web_search_20260209" type`

  - `"web_search" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

  - `?list<string> blockedDomains`

    If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?UserLocation userLocation`

    Parameters for the user's location. Used to provide more relevant search results.

### Web Search Tool 20260318

- `class WebSearchTool20260318`

  - `"web_search_20260318" type`

  - `"web_search" name`

    Name of the tool.

    This is how the tool will be called by the model and in `tool_use` blocks.

  - `?list<AllowedCaller> allowedCallers`

  - `?list<string> allowedDomains`

    If provided, only these domains will be included in results. Cannot be used alongside `blocked_domains`.

  - `?list<string> blockedDomains`

    If provided, these domains will never appear in results. Cannot be used alongside `allowed_domains`.

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?bool deferLoading`

    If true, tool will not be included in initial system prompt. Only loaded when returned via tool_reference from tool search.

  - `?int maxUses`

    Maximum number of times the tool can be used in the API request.

  - `?ResponseInclusion responseInclusion`

    How this tool's result blocks appear in the API response when the result was consumed by a completed code_execution call in the same turn. 'full' returns the complete content (default). 'excluded' drops the nested server_tool_use and result block pair entirely. Results from direct calls, or from code_execution calls that paused before completing, are always returned in full so they can be sent back on the next turn.

  - `?bool strict`

    When true, guarantees schema validation on tool names and inputs

  - `?UserLocation userLocation`

    Parameters for the user's location. Used to provide more relevant search results.

### Web Search Tool Request Error

- `class WebSearchToolRequestError`

  - `"web_search_tool_result_error" type`

  - `WebSearchToolResultErrorCode errorCode`

### Web Search Tool Result Block

- `class WebSearchToolResultBlock`

  - `"web_search_tool_result" type`

  - `Caller caller`

  - `WebSearchToolResultBlockContent content`

  - `string toolUseID`

### Web Search Tool Result Block Content

- `class WebSearchToolResultBlockContent`

  - `class WebSearchToolResultError`

    - `"web_search_tool_result_error" type`

    - `WebSearchToolResultErrorCode errorCode`

  - `class list<WebSearchResultBlock>`

    - `"web_search_result" type`

    - `string encryptedContent`

    - `?string pageAge`

    - `string title`

    - `string url`

### Web Search Tool Result Block Param

- `class WebSearchToolResultBlockParam`

  - `"web_search_tool_result" type`

  - `WebSearchToolResultBlockParamContent content`

  - `string toolUseID`

  - `?CacheControlEphemeral cacheControl`

    Create a cache control breakpoint at this content block.

  - `?Caller caller`

### Web Search Tool Result Block Param Content

- `class WebSearchToolResultBlockParamContent`

  - `class list<WebSearchResultBlockParam>`

    - `"web_search_result" type`

    - `string encryptedContent`

    - `string title`

    - `string url`

    - `?string pageAge`

  - `class WebSearchToolRequestError`

    - `"web_search_tool_result_error" type`

    - `WebSearchToolResultErrorCode errorCode`

### Web Search Tool Result Error

- `class WebSearchToolResultError`

  - `"web_search_tool_result_error" type`

  - `WebSearchToolResultErrorCode errorCode`

### Web Search Tool Result Error Code

- `enum WebSearchToolResultErrorCode`

  - `"invalid_tool_input"`

  - `"unavailable"`

  - `"max_uses_exceeded"`

  - `"too_many_requests"`

  - `"query_too_long"`

  - `"request_too_large"`

## Messages › Batches

### Create a Message Batch

`$client->messages->batches->create(list<Request> requests, ?string userProfileID, ?string workspaceID): MessageBatch`

**POST** `/v1/messages/batches`

Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `requests: list<Request>`

  List of requests for prompt completion. Each is an individual request to create a Message.

- `userProfileID?: optional string` (header parameter)

  The user profile ID to attribute the requests in this batch to. Use when acting on behalf of a party other than your organization. Applies to every request in the batch; an individual request whose `user_profile_id` body field conflicts with this header is errored.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class MessageBatch`

  - `"message_batch" type`

    Object type.

    For Message Batches, this is always `"message_batch"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?\Datetime archivedAt`

    RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.

  - `?\Datetime cancelInitiatedAt`

    RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.

  - `\Datetime createdAt`

    RFC 3339 datetime string representing the time at which the Message Batch was created.

  - `?\Datetime endedAt`

    RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.

    Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.

  - `\Datetime expiresAt`

    RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.

  - `ProcessingStatus processingStatus`

    Processing status of the Message Batch.

  - `MessageBatchRequestCounts requestCounts`

    Tallies requests within the Message Batch, categorized by their status.

    Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.

  - `?string resultsURL`

    URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.

    Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;
use Anthropic\Messages\Model;

$client = new Client(apiKey: 'my-anthropic-api-key');

$messageBatch = $client->messages->batches->create(
  requests: [
    [
      'customID' => 'my-custom-id-1',
      'params' => [
        'maxTokens' => 1024,
        'messages' => [['content' => 'Hello, world', 'role' => 'user']],
        'model' => Model::CLAUDE_OPUS_5_5,
        'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
        'container' => [
          'id' => 'id',
          'skills' => [
            ['skillID' => 'pdf', 'type' => 'anthropic', 'version' => 'latest']
          ],
        ],
        'diagnostics' => ['previousMessageID' => 'previous_message_id'],
        'inferenceGeo' => 'inference_geo',
        'metadata' => ['userID' => '13803d75-b4b5-4c3e-b2a2-6f21399b021b'],
        'outputConfig' => [
          'effort' => 'low',
          'format' => ['schema' => ['foo' => 'bar'], 'type' => 'json_schema'],
        ],
        'serviceTier' => 'auto',
        'speed' => 'standard',
        'stopSequences' => ['string'],
        'stream' => false,
        'system' => [
          [
            'text' => 'Today\'s date is 2024-06-01.',
            'type' => 'text',
            'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
            'citations' => [
              [
                'citedText' => 'The grass is green. The sky is blue.',
                'documentIndex' => 0,
                'documentTitle' => 'x',
                'endCharIndex' => 0,
                'startCharIndex' => 0,
                'type' => 'char_location',
              ],
            ],
          ],
        ],
        'thinking' => ['type' => 'adaptive', 'display' => 'summarized'],
        'toolChoice' => ['type' => 'auto', 'disableParallelToolUse' => true],
        'tools' => [
          [
            'inputSchema' => [
              'type' => 'object',
              'properties' => ['location' => 'bar', 'unit' => 'bar'],
              'required' => ['location'],
            ],
            'name' => 'name',
            'allowedCallers' => ['direct'],
            'cacheControl' => ['type' => 'ephemeral', 'ttl' => '5m'],
            'deferLoading' => true,
            'description' => 'Get the current weather in a given location',
            'eagerInputStreaming' => true,
            'inputExamples' => [['foo' => 'bar']],
            'strict' => true,
            'type' => 'custom',
          ],
        ],
      ],
    ],
  ],
  userProfileID: 'anthropic-user-profile-id',
  workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy',
);

var_dump($messageBatch);
```

##### Response (200)

```json
{
  "id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",
  "archived_at": "2024-08-20T18:37:24.100435Z",
  "cancel_initiated_at": "2024-08-20T18:37:24.100435Z",
  "created_at": "2024-08-20T18:37:24.100435Z",
  "ended_at": "2024-08-20T18:37:24.100435Z",
  "expires_at": "2024-08-20T18:37:24.100435Z",
  "processing_status": "in_progress",
  "request_counts": {
    "canceled": 10,
    "errored": 30,
    "expired": 10,
    "processing": 100,
    "succeeded": 50
  },
  "results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",
  "type": "message_batch"
}
```

### Retrieve a Message Batch

`$client->messages->batches->retrieve(string messageBatchID, ?string workspaceID): MessageBatch`

**GET** `/v1/messages/batches/{message_batch_id}`

This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `messageBatchID: string` (path parameter)

  ID of the Message Batch.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class MessageBatch`

  - `"message_batch" type`

    Object type.

    For Message Batches, this is always `"message_batch"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?\Datetime archivedAt`

    RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.

  - `?\Datetime cancelInitiatedAt`

    RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.

  - `\Datetime createdAt`

    RFC 3339 datetime string representing the time at which the Message Batch was created.

  - `?\Datetime endedAt`

    RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.

    Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.

  - `\Datetime expiresAt`

    RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.

  - `ProcessingStatus processingStatus`

    Processing status of the Message Batch.

  - `MessageBatchRequestCounts requestCounts`

    Tallies requests within the Message Batch, categorized by their status.

    Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.

  - `?string resultsURL`

    URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.

    Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;

$client = new Client(apiKey: 'my-anthropic-api-key');

$messageBatch = $client->messages->batches->retrieve(
  'message_batch_id', workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy'
);

var_dump($messageBatch);
```

##### Response (200)

```json
{
  "id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",
  "archived_at": "2024-08-20T18:37:24.100435Z",
  "cancel_initiated_at": "2024-08-20T18:37:24.100435Z",
  "created_at": "2024-08-20T18:37:24.100435Z",
  "ended_at": "2024-08-20T18:37:24.100435Z",
  "expires_at": "2024-08-20T18:37:24.100435Z",
  "processing_status": "in_progress",
  "request_counts": {
    "canceled": 10,
    "errored": 30,
    "expired": 10,
    "processing": 100,
    "succeeded": 50
  },
  "results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",
  "type": "message_batch"
}
```

### List Message Batches

`$client->messages->batches->list(?string afterID, ?string beforeID, ?int limit, ?string workspaceID): Page<MessageBatch>`

**GET** `/v1/messages/batches`

List all Message Batches within a Workspace. Most recently created batches are returned first.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `afterID?: optional string` (query parameter)

  ID of the object to use as a cursor for pagination. When provided, returns the page of results immediately after this object.

- `beforeID?: optional string` (query parameter)

  ID of the object to use as a cursor for pagination. When provided, returns the page of results immediately before this object.

- `limit?: optional int` (query parameter)

  Number of items to return per page.

  Defaults to `20`. Ranges from `1` to `1000`.

  default: 20

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class MessageBatch`

  - `"message_batch" type`

    Object type.

    For Message Batches, this is always `"message_batch"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?\Datetime archivedAt`

    RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.

  - `?\Datetime cancelInitiatedAt`

    RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.

  - `\Datetime createdAt`

    RFC 3339 datetime string representing the time at which the Message Batch was created.

  - `?\Datetime endedAt`

    RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.

    Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.

  - `\Datetime expiresAt`

    RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.

  - `ProcessingStatus processingStatus`

    Processing status of the Message Batch.

  - `MessageBatchRequestCounts requestCounts`

    Tallies requests within the Message Batch, categorized by their status.

    Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.

  - `?string resultsURL`

    URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.

    Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;

$client = new Client(apiKey: 'my-anthropic-api-key');

$page = $client->messages->batches->list(
  afterID: 'after_id',
  beforeID: 'before_id',
  limit: 1,
  workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy',
);

var_dump($page);
```

##### Response (200)

```json
{
  "data": [
    {
      "id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",
      "archived_at": "2024-08-20T18:37:24.100435Z",
      "cancel_initiated_at": "2024-08-20T18:37:24.100435Z",
      "created_at": "2024-08-20T18:37:24.100435Z",
      "ended_at": "2024-08-20T18:37:24.100435Z",
      "expires_at": "2024-08-20T18:37:24.100435Z",
      "processing_status": "in_progress",
      "request_counts": {
        "canceled": 10,
        "errored": 30,
        "expired": 10,
        "processing": 100,
        "succeeded": 50
      },
      "results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",
      "type": "message_batch"
    }
  ],
  "first_id": "first_id",
  "has_more": true,
  "last_id": "last_id"
}
```

### Cancel a Message Batch

`$client->messages->batches->cancel(string messageBatchID, ?string workspaceID): MessageBatch`

**POST** `/v1/messages/batches/{message_batch_id}/cancel`

Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `messageBatchID: string` (path parameter)

  ID of the Message Batch.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class MessageBatch`

  - `"message_batch" type`

    Object type.

    For Message Batches, this is always `"message_batch"`.

  - `string id`

    Unique object identifier.

    The format and length of IDs may change over time.

  - `?\Datetime archivedAt`

    RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.

  - `?\Datetime cancelInitiatedAt`

    RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.

  - `\Datetime createdAt`

    RFC 3339 datetime string representing the time at which the Message Batch was created.

  - `?\Datetime endedAt`

    RFC 3339 datetime string representing the time at which processing for the Message Batch ended. Specified only once processing ends.

    Processing ends when every request in a Message Batch has either succeeded, errored, canceled, or expired.

  - `\Datetime expiresAt`

    RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.

  - `ProcessingStatus processingStatus`

    Processing status of the Message Batch.

  - `MessageBatchRequestCounts requestCounts`

    Tallies requests within the Message Batch, categorized by their status.

    Requests start as `processing` and move to one of the other statuses only once processing of the entire batch ends. The sum of all values always matches the total number of requests in the batch.

  - `?string resultsURL`

    URL to a `.jsonl` file containing the results of the Message Batch requests. Specified only once processing ends.

    Results in the file are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;

$client = new Client(apiKey: 'my-anthropic-api-key');

$messageBatch = $client->messages->batches->cancel(
  'message_batch_id', workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy'
);

var_dump($messageBatch);
```

##### Response (200)

```json
{
  "id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",
  "archived_at": "2024-08-20T18:37:24.100435Z",
  "cancel_initiated_at": "2024-08-20T18:37:24.100435Z",
  "created_at": "2024-08-20T18:37:24.100435Z",
  "ended_at": "2024-08-20T18:37:24.100435Z",
  "expires_at": "2024-08-20T18:37:24.100435Z",
  "processing_status": "in_progress",
  "request_counts": {
    "canceled": 10,
    "errored": 30,
    "expired": 10,
    "processing": 100,
    "succeeded": 50
  },
  "results_url": "https://api.anthropic.com/v1/messages/batches/msgbatch_013Zva2CMHLNnXjNJJKqJ2EF/results",
  "type": "message_batch"
}
```

### Delete a Message Batch

`$client->messages->batches->delete(string messageBatchID, ?string workspaceID): DeletedMessageBatch`

**DELETE** `/v1/messages/batches/{message_batch_id}`

Delete a Message Batch.

Message Batches can only be deleted once they've finished processing. If you'd like to delete an in-progress batch, you must first cancel it.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `messageBatchID: string` (path parameter)

  ID of the Message Batch.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class DeletedMessageBatch`

  - `"message_batch_deleted" type`

    Deleted object type.

    For Message Batches, this is always `"message_batch_deleted"`.

  - `string id`

    ID of the Message Batch.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;

$client = new Client(apiKey: 'my-anthropic-api-key');

$deletedMessageBatch = $client->messages->batches->delete(
  'message_batch_id', workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy'
);

var_dump($deletedMessageBatch);
```

##### Response (200)

```json
{
  "id": "msgbatch_013Zva2CMHLNnXjNJJKqJ2EF",
  "type": "message_batch_deleted"
}
```

### Retrieve Message Batch results

`$client->messages->batches->resultsStream(string messageBatchID, ?string workspaceID): BaseStream<MessageBatchIndividualResponse>`

**GET** `/v1/messages/batches/{message_batch_id}/results`

Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.

Learn more about the Message Batches API in our [user guide](https://platform.claude.com/docs/en/build-with-claude/batch-processing?sdk_languages=php)

#### Parameters

- `messageBatchID: string` (path parameter)

  ID of the Message Batch.

- `workspaceID?: optional string` (header parameter)

  Optional header to select the Workspace for this request. The value is a Workspace ID (for example, `wrkspc_011CZkZaBF1tNoB5wlCeusgy`).

  Only needed for credentials that can act on more than one Workspace. A credential that belongs to a specific Workspace may omit it; if sent, it must match that Workspace.

#### Returns

- `class MessageBatchIndividualResponse`

  - `string customID`

    Developer-provided ID created for each request in a Message Batch. Useful for matching results to requests, as results may be given out of request order.

    Must be unique for each request within the Message Batch.

  - `MessageBatchResult result`

    Processing result for this request.

    Contains a Message output if processing was successful, an error response if processing failed, or the reason why processing was not attempted, such as cancellation or expiration.

#### Example

```php
<?php

require_once dirname(__DIR__) . '/vendor/autoload.php';

use Anthropic\Client;

$client = new Client(apiKey: 'my-anthropic-api-key');

$messageBatchIndividualResponse = $client->messages->batches->resultsStream(
  'message_batch_id', workspaceID: 'wrkspc_011CZkZaBF1tNoB5wlCeusgy'
);

var_dump($messageBatchIndividualResponse);
```
