Batches

class DeletedMessageBatch:

String id

ID of the Message Batch.

JsonValue; type

Deleted object type.

For Message Batches, this is always "message_batch_deleted".

Accepts one of the following:

MESSAGE_BATCH_DELETED("message_batch_deleted")

class MessageBatch:

String id

Unique object identifier.

The format and length of IDs may change over time.

Optional<LocalDateTime> archivedAt

RFC 3339 datetime string representing the time at which the Message Batch was archived and its results became unavailable.

formatdate-time

Optional<LocalDateTime> cancelInitiatedAt

RFC 3339 datetime string representing the time at which cancellation was initiated for the Message Batch. Specified only if cancellation was initiated.

formatdate-time

LocalDateTime createdAt

RFC 3339 datetime string representing the time at which the Message Batch was created.

formatdate-time

Optional<LocalDateTime> 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.

formatdate-time

LocalDateTime expiresAt

RFC 3339 datetime string representing the time at which the Message Batch will expire and end processing, which is 24 hours after creation.

formatdate-time

ProcessingStatus processingStatus

Processing status of the Message Batch.

Accepts one of the following:

IN_PROGRESS("in_progress")

CANCELING("canceling")

ENDED("ended")

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.

long canceled

Number of requests in the Message Batch that have been canceled.

This is zero until processing of the entire Message Batch has ended.

long errored

Number of requests in the Message Batch that encountered an error.

This is zero until processing of the entire Message Batch has ended.

long expired

Number of requests in the Message Batch that have expired.

This is zero until processing of the entire Message Batch has ended.

long processing

Number of requests in the Message Batch that are processing.

long succeeded

Number of requests in the Message Batch that have completed successfully.

This is zero until processing of the entire Message Batch has ended.

Optional<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.

JsonValue; type

Object type.

For Message Batches, this is always "message_batch".

Accepts one of the following:

MESSAGE_BATCH("message_batch")

class MessageBatchCanceledResult:

JsonValue; type

Accepts one of the following:

CANCELED("canceled")

class MessageBatchErroredResult:

ErrorResponse error

ErrorObject error

Accepts one of the following:

class InvalidRequestError:

String message

JsonValue; type

Accepts one of the following:

INVALID_REQUEST_ERROR("invalid_request_error")

class AuthenticationError:

String message

JsonValue; type

Accepts one of the following:

AUTHENTICATION_ERROR("authentication_error")

class BillingError:

String message

JsonValue; type

Accepts one of the following:

BILLING_ERROR("billing_error")

class PermissionError:

String message

JsonValue; type

Accepts one of the following:

PERMISSION_ERROR("permission_error")

class NotFoundError:

String message

JsonValue; type

Accepts one of the following:

NOT_FOUND_ERROR("not_found_error")

class RateLimitError:

String message

JsonValue; type

Accepts one of the following:

RATE_LIMIT_ERROR("rate_limit_error")

class GatewayTimeoutError:

String message

JsonValue; type

Accepts one of the following:

TIMEOUT_ERROR("timeout_error")

class ApiErrorObject:

String message

JsonValue; type

Accepts one of the following:

API_ERROR("api_error")

class OverloadedError:

String message

JsonValue; type

Accepts one of the following:

OVERLOADED_ERROR("overloaded_error")

Optional<String> requestId

JsonValue; type

Accepts one of the following:

ERROR("error")

JsonValue; type

Accepts one of the following:

ERRORED("errored")

class MessageBatchExpiredResult:

JsonValue; type

Accepts one of the following:

EXPIRED("expired")

class MessageBatchIndividualResponse:

This is a single line in the response .jsonl file and does not represent the response as a whole.

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.

Accepts one of the following:

class MessageBatchSucceededResult:

Message message

String id

Unique object identifier.

The format and length of IDs may change over time.

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:

[{"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:

[
  {"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:

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

Accepts one of the following:

class TextBlock:

Optional<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.

Accepts one of the following:

class CitationCharLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endCharIndex

Optional<String> fileId

long startCharIndex

JsonValue; type

Accepts one of the following:

CHAR_LOCATION("char_location")

class CitationPageLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endPageNumber

Optional<String> fileId

long startPageNumber

JsonValue; type

Accepts one of the following:

PAGE_LOCATION("page_location")

class CitationContentBlockLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endBlockIndex

Optional<String> fileId

long startBlockIndex

JsonValue; type

Accepts one of the following:

CONTENT_BLOCK_LOCATION("content_block_location")

class CitationsWebSearchResultLocation:

String citedText

String encryptedIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT_LOCATION("web_search_result_location")

String url

class CitationsSearchResultLocation:

String citedText

long endBlockIndex

long searchResultIndex

String source

long startBlockIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

SEARCH_RESULT_LOCATION("search_result_location")

String text

JsonValue; type

Accepts one of the following:

TEXT("text")

class ThinkingBlock:

String signature

String thinking

JsonValue; type

Accepts one of the following:

THINKING("thinking")

class RedactedThinkingBlock:

String data

JsonValue; type

Accepts one of the following:

REDACTED_THINKING("redacted_thinking")

class ToolUseBlock:

String id

Input input

String name

JsonValue; type

Accepts one of the following:

TOOL_USE("tool_use")

class ServerToolUseBlock:

String id

Input input

JsonValue; name

Accepts one of the following:

WEB_SEARCH("web_search")

JsonValue; type

Accepts one of the following:

SERVER_TOOL_USE("server_tool_use")

class WebSearchToolResultBlock:

WebSearchToolResultBlockContent content

Accepts one of the following:

class WebSearchToolResultError:

ErrorCode errorCode

Accepts one of the following:

INVALID_TOOL_INPUT("invalid_tool_input")

UNAVAILABLE("unavailable")

MAX_USES_EXCEEDED("max_uses_exceeded")

TOO_MANY_REQUESTS("too_many_requests")

QUERY_TOO_LONG("query_too_long")

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT_ERROR("web_search_tool_result_error")

List<WebSearchResultBlock>

String encryptedContent

Optional<String> pageAge

String title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT("web_search_result")

String url

String toolUseId

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT("web_search_tool_result")

Model model

The model that will complete your prompt.

See models for additional details and options.

Accepts one of the following:

CLAUDE_OPUS_4_5_20251101("claude-opus-4-5-20251101")

Premium model combining maximum intelligence with practical performance

CLAUDE_OPUS_4_5("claude-opus-4-5")

Premium model combining maximum intelligence with practical performance

CLAUDE_3_7_SONNET_LATEST("claude-3-7-sonnet-latest")

High-performance model with early extended thinking

CLAUDE_3_7_SONNET_20250219("claude-3-7-sonnet-20250219")

High-performance model with early extended thinking

CLAUDE_3_5_HAIKU_LATEST("claude-3-5-haiku-latest")

Fastest and most compact model for near-instant responsiveness

CLAUDE_3_5_HAIKU_20241022("claude-3-5-haiku-20241022")

Our fastest model

CLAUDE_HAIKU_4_5("claude-haiku-4-5")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_HAIKU_4_5_20251001("claude-haiku-4-5-20251001")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_SONNET_4_20250514("claude-sonnet-4-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_0("claude-sonnet-4-0")

High-performance model with extended thinking

CLAUDE_4_SONNET_20250514("claude-4-sonnet-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_5("claude-sonnet-4-5")

Our best model for real-world agents and coding

CLAUDE_SONNET_4_5_20250929("claude-sonnet-4-5-20250929")

Our best model for real-world agents and coding

CLAUDE_OPUS_4_0("claude-opus-4-0")

Our most capable model

CLAUDE_OPUS_4_20250514("claude-opus-4-20250514")

Our most capable model

CLAUDE_4_OPUS_20250514("claude-4-opus-20250514")

Our most capable model

CLAUDE_OPUS_4_1_20250805("claude-opus-4-1-20250805")

Our most capable model

CLAUDE_3_OPUS_LATEST("claude-3-opus-latest")

Excels at writing and complex tasks

CLAUDE_3_OPUS_20240229("claude-3-opus-20240229")

Excels at writing and complex tasks

CLAUDE_3_HAIKU_20240307("claude-3-haiku-20240307")

Our previous most fast and cost-effective

JsonValue; role

Conversational role of the generated message.

This will always be "assistant".

Accepts one of the following:

ASSISTANT("assistant")

Optional<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

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.

Accepts one of the following:

END_TURN("end_turn")

MAX_TOKENS("max_tokens")

STOP_SEQUENCE("stop_sequence")

TOOL_USE("tool_use")

PAUSE_TURN("pause_turn")

REFUSAL("refusal")

Optional<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.

JsonValue; type

Object type.

For Messages, this is always "message".

Accepts one of the following:

MESSAGE("message")

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.

Optional<CacheCreation> cacheCreation

Breakdown of cached tokens by TTL

long ephemeral1hInputTokens

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

minimum0

long ephemeral5mInputTokens

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

minimum0

Optional<Long> cacheCreationInputTokens

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

minimum0

Optional<Long> cacheReadInputTokens

The number of input tokens read from the cache.

minimum0

long inputTokens

The number of input tokens which were used.

minimum0

long outputTokens

The number of output tokens which were used.

minimum0

Optional<ServerToolUsage> serverToolUse

The number of server tool requests.

long webSearchRequests

The number of web search tool requests.

minimum0

Optional<ServiceTier> serviceTier

If the request used the priority, standard, or batch tier.

Accepts one of the following:

STANDARD("standard")

PRIORITY("priority")

BATCH("batch")

JsonValue; type

Accepts one of the following:

SUCCEEDED("succeeded")

class MessageBatchErroredResult:

ErrorResponse error

ErrorObject error

Accepts one of the following:

class InvalidRequestError:

String message

JsonValue; type

Accepts one of the following:

INVALID_REQUEST_ERROR("invalid_request_error")

class AuthenticationError:

String message

JsonValue; type

Accepts one of the following:

AUTHENTICATION_ERROR("authentication_error")

class BillingError:

String message

JsonValue; type

Accepts one of the following:

BILLING_ERROR("billing_error")

class PermissionError:

String message

JsonValue; type

Accepts one of the following:

PERMISSION_ERROR("permission_error")

class NotFoundError:

String message

JsonValue; type

Accepts one of the following:

NOT_FOUND_ERROR("not_found_error")

class RateLimitError:

String message

JsonValue; type

Accepts one of the following:

RATE_LIMIT_ERROR("rate_limit_error")

class GatewayTimeoutError:

String message

JsonValue; type

Accepts one of the following:

TIMEOUT_ERROR("timeout_error")

class ApiErrorObject:

String message

JsonValue; type

Accepts one of the following:

API_ERROR("api_error")

class OverloadedError:

String message

JsonValue; type

Accepts one of the following:

OVERLOADED_ERROR("overloaded_error")

Optional<String> requestId

JsonValue; type

Accepts one of the following:

ERROR("error")

JsonValue; type

Accepts one of the following:

ERRORED("errored")

class MessageBatchCanceledResult:

JsonValue; type

Accepts one of the following:

CANCELED("canceled")

class MessageBatchExpiredResult:

JsonValue; type

Accepts one of the following:

EXPIRED("expired")

class MessageBatchRequestCounts:

long canceled

Number of requests in the Message Batch that have been canceled.

This is zero until processing of the entire Message Batch has ended.

long errored

Number of requests in the Message Batch that encountered an error.

This is zero until processing of the entire Message Batch has ended.

long expired

Number of requests in the Message Batch that have expired.

This is zero until processing of the entire Message Batch has ended.

long processing

Number of requests in the Message Batch that are processing.

long succeeded

Number of requests in the Message Batch that have completed successfully.

This is zero until processing of the entire Message Batch has ended.

class MessageBatchResult:

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.

class MessageBatchSucceededResult:

Message message

String id

Unique object identifier.

The format and length of IDs may change over time.

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:

[{"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:

[
  {"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:

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

Accepts one of the following:

class TextBlock:

Optional<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.

Accepts one of the following:

class CitationCharLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endCharIndex

Optional<String> fileId

long startCharIndex

JsonValue; type

Accepts one of the following:

CHAR_LOCATION("char_location")

class CitationPageLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endPageNumber

Optional<String> fileId

long startPageNumber

JsonValue; type

Accepts one of the following:

PAGE_LOCATION("page_location")

class CitationContentBlockLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endBlockIndex

Optional<String> fileId

long startBlockIndex

JsonValue; type

Accepts one of the following:

CONTENT_BLOCK_LOCATION("content_block_location")

class CitationsWebSearchResultLocation:

String citedText

String encryptedIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT_LOCATION("web_search_result_location")

String url

class CitationsSearchResultLocation:

String citedText

long endBlockIndex

long searchResultIndex

String source

long startBlockIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

SEARCH_RESULT_LOCATION("search_result_location")

String text

JsonValue; type

Accepts one of the following:

TEXT("text")

class ThinkingBlock:

String signature

String thinking

JsonValue; type

Accepts one of the following:

THINKING("thinking")

class RedactedThinkingBlock:

String data

JsonValue; type

Accepts one of the following:

REDACTED_THINKING("redacted_thinking")

class ToolUseBlock:

String id

Input input

String name

JsonValue; type

Accepts one of the following:

TOOL_USE("tool_use")

class ServerToolUseBlock:

String id

Input input

JsonValue; name

Accepts one of the following:

WEB_SEARCH("web_search")

JsonValue; type

Accepts one of the following:

SERVER_TOOL_USE("server_tool_use")

class WebSearchToolResultBlock:

WebSearchToolResultBlockContent content

Accepts one of the following:

class WebSearchToolResultError:

ErrorCode errorCode

Accepts one of the following:

INVALID_TOOL_INPUT("invalid_tool_input")

UNAVAILABLE("unavailable")

MAX_USES_EXCEEDED("max_uses_exceeded")

TOO_MANY_REQUESTS("too_many_requests")

QUERY_TOO_LONG("query_too_long")

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT_ERROR("web_search_tool_result_error")

List<WebSearchResultBlock>

String encryptedContent

Optional<String> pageAge

String title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT("web_search_result")

String url

String toolUseId

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT("web_search_tool_result")

Model model

The model that will complete your prompt.

See models for additional details and options.

Accepts one of the following:

CLAUDE_OPUS_4_5_20251101("claude-opus-4-5-20251101")

Premium model combining maximum intelligence with practical performance

CLAUDE_OPUS_4_5("claude-opus-4-5")

Premium model combining maximum intelligence with practical performance

CLAUDE_3_7_SONNET_LATEST("claude-3-7-sonnet-latest")

High-performance model with early extended thinking

CLAUDE_3_7_SONNET_20250219("claude-3-7-sonnet-20250219")

High-performance model with early extended thinking

CLAUDE_3_5_HAIKU_LATEST("claude-3-5-haiku-latest")

Fastest and most compact model for near-instant responsiveness

CLAUDE_3_5_HAIKU_20241022("claude-3-5-haiku-20241022")

Our fastest model

CLAUDE_HAIKU_4_5("claude-haiku-4-5")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_HAIKU_4_5_20251001("claude-haiku-4-5-20251001")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_SONNET_4_20250514("claude-sonnet-4-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_0("claude-sonnet-4-0")

High-performance model with extended thinking

CLAUDE_4_SONNET_20250514("claude-4-sonnet-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_5("claude-sonnet-4-5")

Our best model for real-world agents and coding

CLAUDE_SONNET_4_5_20250929("claude-sonnet-4-5-20250929")

Our best model for real-world agents and coding

CLAUDE_OPUS_4_0("claude-opus-4-0")

Our most capable model

CLAUDE_OPUS_4_20250514("claude-opus-4-20250514")

Our most capable model

CLAUDE_4_OPUS_20250514("claude-4-opus-20250514")

Our most capable model

CLAUDE_OPUS_4_1_20250805("claude-opus-4-1-20250805")

Our most capable model

CLAUDE_3_OPUS_LATEST("claude-3-opus-latest")

Excels at writing and complex tasks

CLAUDE_3_OPUS_20240229("claude-3-opus-20240229")

Excels at writing and complex tasks

CLAUDE_3_HAIKU_20240307("claude-3-haiku-20240307")

Our previous most fast and cost-effective

JsonValue; role

Conversational role of the generated message.

This will always be "assistant".

Accepts one of the following:

ASSISTANT("assistant")

Optional<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

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.

Accepts one of the following:

END_TURN("end_turn")

MAX_TOKENS("max_tokens")

STOP_SEQUENCE("stop_sequence")

TOOL_USE("tool_use")

PAUSE_TURN("pause_turn")

REFUSAL("refusal")

Optional<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.

JsonValue; type

Object type.

For Messages, this is always "message".

Accepts one of the following:

MESSAGE("message")

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.

Optional<CacheCreation> cacheCreation

Breakdown of cached tokens by TTL

long ephemeral1hInputTokens

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

minimum0

long ephemeral5mInputTokens

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

minimum0

Optional<Long> cacheCreationInputTokens

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

minimum0

Optional<Long> cacheReadInputTokens

The number of input tokens read from the cache.

minimum0

long inputTokens

The number of input tokens which were used.

minimum0

long outputTokens

The number of output tokens which were used.

minimum0

Optional<ServerToolUsage> serverToolUse

The number of server tool requests.

long webSearchRequests

The number of web search tool requests.

minimum0

Optional<ServiceTier> serviceTier

If the request used the priority, standard, or batch tier.

Accepts one of the following:

STANDARD("standard")

PRIORITY("priority")

BATCH("batch")

JsonValue; type

Accepts one of the following:

SUCCEEDED("succeeded")

class MessageBatchErroredResult:

ErrorResponse error

ErrorObject error

Accepts one of the following:

class InvalidRequestError:

String message

JsonValue; type

Accepts one of the following:

INVALID_REQUEST_ERROR("invalid_request_error")

class AuthenticationError:

String message

JsonValue; type

Accepts one of the following:

AUTHENTICATION_ERROR("authentication_error")

class BillingError:

String message

JsonValue; type

Accepts one of the following:

BILLING_ERROR("billing_error")

class PermissionError:

String message

JsonValue; type

Accepts one of the following:

PERMISSION_ERROR("permission_error")

class NotFoundError:

String message

JsonValue; type

Accepts one of the following:

NOT_FOUND_ERROR("not_found_error")

class RateLimitError:

String message

JsonValue; type

Accepts one of the following:

RATE_LIMIT_ERROR("rate_limit_error")

class GatewayTimeoutError:

String message

JsonValue; type

Accepts one of the following:

TIMEOUT_ERROR("timeout_error")

class ApiErrorObject:

String message

JsonValue; type

Accepts one of the following:

API_ERROR("api_error")

class OverloadedError:

String message

JsonValue; type

Accepts one of the following:

OVERLOADED_ERROR("overloaded_error")

Optional<String> requestId

JsonValue; type

Accepts one of the following:

ERROR("error")

JsonValue; type

Accepts one of the following:

ERRORED("errored")

class MessageBatchCanceledResult:

JsonValue; type

Accepts one of the following:

CANCELED("canceled")

class MessageBatchExpiredResult:

JsonValue; type

Accepts one of the following:

EXPIRED("expired")

class MessageBatchSucceededResult:

Message message

String id

Unique object identifier.

The format and length of IDs may change over time.

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:

[{"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:

[
  {"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:

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

Accepts one of the following:

class TextBlock:

Optional<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.

Accepts one of the following:

class CitationCharLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endCharIndex

Optional<String> fileId

long startCharIndex

JsonValue; type

Accepts one of the following:

CHAR_LOCATION("char_location")

class CitationPageLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endPageNumber

Optional<String> fileId

long startPageNumber

JsonValue; type

Accepts one of the following:

PAGE_LOCATION("page_location")

class CitationContentBlockLocation:

String citedText

long documentIndex

Optional<String> documentTitle

long endBlockIndex

Optional<String> fileId

long startBlockIndex

JsonValue; type

Accepts one of the following:

CONTENT_BLOCK_LOCATION("content_block_location")

class CitationsWebSearchResultLocation:

String citedText

String encryptedIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT_LOCATION("web_search_result_location")

String url

class CitationsSearchResultLocation:

String citedText

long endBlockIndex

long searchResultIndex

String source

long startBlockIndex

Optional<String> title

JsonValue; type

Accepts one of the following:

SEARCH_RESULT_LOCATION("search_result_location")

String text

JsonValue; type

Accepts one of the following:

TEXT("text")

class ThinkingBlock:

String signature

String thinking

JsonValue; type

Accepts one of the following:

THINKING("thinking")

class RedactedThinkingBlock:

String data

JsonValue; type

Accepts one of the following:

REDACTED_THINKING("redacted_thinking")

class ToolUseBlock:

String id

Input input

String name

JsonValue; type

Accepts one of the following:

TOOL_USE("tool_use")

class ServerToolUseBlock:

String id

Input input

JsonValue; name

Accepts one of the following:

WEB_SEARCH("web_search")

JsonValue; type

Accepts one of the following:

SERVER_TOOL_USE("server_tool_use")

class WebSearchToolResultBlock:

WebSearchToolResultBlockContent content

Accepts one of the following:

class WebSearchToolResultError:

ErrorCode errorCode

Accepts one of the following:

INVALID_TOOL_INPUT("invalid_tool_input")

UNAVAILABLE("unavailable")

MAX_USES_EXCEEDED("max_uses_exceeded")

TOO_MANY_REQUESTS("too_many_requests")

QUERY_TOO_LONG("query_too_long")

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT_ERROR("web_search_tool_result_error")

List<WebSearchResultBlock>

String encryptedContent

Optional<String> pageAge

String title

JsonValue; type

Accepts one of the following:

WEB_SEARCH_RESULT("web_search_result")

String url

String toolUseId

JsonValue; type

Accepts one of the following:

WEB_SEARCH_TOOL_RESULT("web_search_tool_result")

Model model

The model that will complete your prompt.

See models for additional details and options.

Accepts one of the following:

CLAUDE_OPUS_4_5_20251101("claude-opus-4-5-20251101")

Premium model combining maximum intelligence with practical performance

CLAUDE_OPUS_4_5("claude-opus-4-5")

Premium model combining maximum intelligence with practical performance

CLAUDE_3_7_SONNET_LATEST("claude-3-7-sonnet-latest")

High-performance model with early extended thinking

CLAUDE_3_7_SONNET_20250219("claude-3-7-sonnet-20250219")

High-performance model with early extended thinking

CLAUDE_3_5_HAIKU_LATEST("claude-3-5-haiku-latest")

Fastest and most compact model for near-instant responsiveness

CLAUDE_3_5_HAIKU_20241022("claude-3-5-haiku-20241022")

Our fastest model

CLAUDE_HAIKU_4_5("claude-haiku-4-5")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_HAIKU_4_5_20251001("claude-haiku-4-5-20251001")

Hybrid model, capable of near-instant responses and extended thinking

CLAUDE_SONNET_4_20250514("claude-sonnet-4-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_0("claude-sonnet-4-0")

High-performance model with extended thinking

CLAUDE_4_SONNET_20250514("claude-4-sonnet-20250514")

High-performance model with extended thinking

CLAUDE_SONNET_4_5("claude-sonnet-4-5")

Our best model for real-world agents and coding

CLAUDE_SONNET_4_5_20250929("claude-sonnet-4-5-20250929")

Our best model for real-world agents and coding

CLAUDE_OPUS_4_0("claude-opus-4-0")

Our most capable model

CLAUDE_OPUS_4_20250514("claude-opus-4-20250514")

Our most capable model

CLAUDE_4_OPUS_20250514("claude-4-opus-20250514")

Our most capable model

CLAUDE_OPUS_4_1_20250805("claude-opus-4-1-20250805")

Our most capable model

CLAUDE_3_OPUS_LATEST("claude-3-opus-latest")

Excels at writing and complex tasks

CLAUDE_3_OPUS_20240229("claude-3-opus-20240229")

Excels at writing and complex tasks

CLAUDE_3_HAIKU_20240307("claude-3-haiku-20240307")

Our previous most fast and cost-effective

JsonValue; role

Conversational role of the generated message.

This will always be "assistant".

Accepts one of the following:

ASSISTANT("assistant")

Optional<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

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.

Accepts one of the following:

END_TURN("end_turn")

MAX_TOKENS("max_tokens")

STOP_SEQUENCE("stop_sequence")

TOOL_USE("tool_use")

PAUSE_TURN("pause_turn")

REFUSAL("refusal")

Optional<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.

JsonValue; type

Object type.

For Messages, this is always "message".

Accepts one of the following:

MESSAGE("message")

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.

Optional<CacheCreation> cacheCreation

Breakdown of cached tokens by TTL

long ephemeral1hInputTokens

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

minimum0

long ephemeral5mInputTokens

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

minimum0

Optional<Long> cacheCreationInputTokens

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

minimum0

Optional<Long> cacheReadInputTokens

The number of input tokens read from the cache.

minimum0

long inputTokens

The number of input tokens which were used.

minimum0

long outputTokens

The number of output tokens which were used.

minimum0

Optional<ServerToolUsage> serverToolUse

The number of server tool requests.

long webSearchRequests

The number of web search tool requests.

minimum0

Optional<ServiceTier> serviceTier

If the request used the priority, standard, or batch tier.

Accepts one of the following:

STANDARD("standard")

PRIORITY("priority")

BATCH("batch")

JsonValue; type

Accepts one of the following:

SUCCEEDED("succeeded")

Create a Message Batch

Retrieve a Message Batch

List Message Batches

Cancel a Message Batch

Delete a Message Batch

Retrieve Message Batch results

ModelsExpand Collapse