Recuperar resultados de um lote de mensagens
Faz streaming dos resultados de um lote de mensagens como um arquivo .jsonl.
Cada linha do arquivo é um objeto JSON contendo o resultado de uma única solicitação no lote de mensagens. Não há garantia de que os resultados estejam na mesma ordem das solicitações. Use o campo custom_id para corresponder resultados a solicitações.
Saiba mais sobre a API Message Batches em nosso guia do usuário
Parameters
ID of the Message Batch.
Returns
message_batch_individual_response: object{ custom_id, result }This is a single line in the response .jsonl file and does not represent the response as a whole.
This is a single line in the response .jsonl file and does not represent the response as a whole.
custom_id: stringDeveloper-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.
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.
result: MessageBatchSucceededResult { message, type } or MessageBatchErroredResult { error, type } or MessageBatchCanceledResult { type } or MessageBatchExpiredResult { type }Processing result for this request.
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.
message_batch_succeeded_result: object{ message, type }
message: object{ id, container, content, 7 more }
id: stringUnique object identifier.
Unique object identifier.
The format and length of IDs may change over time.
container: object{ id, expires_at, skills }Information about the container used in the request (for the code execution tool)
Information about the container used in the request (for the code execution tool)
Identifier for the container used in this request
expires_at: stringThe time at which the container will expire.
The time at which the container will expire.
skills: array of ContainerSkill { skill_id, type, version }Skills loaded in the container
Skills loaded in the container
skill_id: stringSkill ID
Skill ID
type: "anthropic" or "custom"Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)
Type of skill - either 'anthropic' (built-in) or 'custom' (user-defined)
version: stringThe resolved version: a skill version ID for custom skills.
The resolved version: a skill version ID for custom skills.
content: array of ContentBlockContent generated by the model.
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)"}]text_block: object{ citations, text, type }
citations: array of TextCitationCitations supporting the text block.
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.
citation_char_location: object{ cited_text, document_index, document_title, 4 more }
document_index: number
start_char_index: number
citation_page_location: object{ cited_text, document_index, document_title, 4 more }
document_index: number
start_page_number: number
citation_content_block_location: object{ cited_text, document_index, document_title, 4 more }
cited_text: stringThe full text of the cited block range, concatenated.
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.
document_index: number
end_block_index: numberExclusive 0-based end index of the cited block range in the source's content array.
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.
start_block_index: number0-based index of the first cited block in the source's content array.
0-based index of the first cited block in the source's content array.
citations_web_search_result_location: object{ cited_text, encrypted_index, title, 2 more }
title: string
citations_search_result_location: object{ cited_text, end_block_index, search_result_index, 4 more }
cited_text: stringThe full text of the cited block range, concatenated.
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.
end_block_index: numberExclusive 0-based end index of the cited block range in the source's content array.
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.
search_result_index: number0-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.
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.
start_block_index: number0-based index of the first cited block in the source's content array.
0-based index of the first cited block in the source's content array.
text: string
thinking_block: object{ signature, thinking, type }
signature: stringA value used to verify that this thinking block was generated by Claude when it is passed back to the API.
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 for details.
The text of Claude's thinking process for this block.
redacted_thinking_block: object{ data, type }
data: stringThe 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.
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 for details.
tool_use_block: object{ id, caller, input, 3 more }
id: string
caller: DirectCaller { type } or ServerToolCaller { tool_id, type } or ServerToolCaller20260120 { tool_id, type }Tool invocation directly from the model.
Tool invocation directly from the model.
direct_caller: object{ type }Tool invocation directly from the model.
Tool invocation directly from the model.
server_tool_caller: object{ tool_id, type }Tool invocation generated by a server-side tool.
Tool invocation generated by a server-side tool.
tool_id: string
server_tool_caller_20260120: object{ tool_id, type }
tool_id: string
name: string
toolset_name: optional stringFor a toolset member tool_use, the toolset family.
For a toolset member tool_use, the toolset family.
server_tool_use_block: object{ id, caller, input, 2 more }
id: string
caller: DirectCaller { type } or ServerToolCaller { tool_id, type } or ServerToolCaller20260120 { tool_id, type }Tool invocation directly from the model.
Tool invocation directly from the model.
direct_caller: object{ type }Tool invocation directly from the model.
Tool invocation directly from the model.
server_tool_caller: object{ tool_id, type }Tool invocation generated by a server-side tool.
Tool invocation generated by a server-side tool.
tool_id: string
server_tool_caller_20260120: object{ tool_id, type }
tool_id: string
name: "web_search" or "web_fetch" or "code_execution" or 4 more
web_search_tool_result_block: object{ caller, content, tool_use_id, type }
caller: DirectCaller { type } or ServerToolCaller { tool_id, type } or ServerToolCaller20260120 { tool_id, type }Tool invocation directly from the model.
Tool invocation directly from the model.
direct_caller: object{ type }Tool invocation directly from the model.
Tool invocation directly from the model.
server_tool_caller: object{ tool_id, type }Tool invocation generated by a server-side tool.
Tool invocation generated by a server-side tool.
tool_id: string
server_tool_caller_20260120: object{ tool_id, type }
tool_id: string
content: WebSearchToolResultError { error_code, type } or array of WebSearchResultBlock { encrypted_content, page_age, title, 2 more }
web_search_tool_result_error: object{ error_code, type }
error_code: "invalid_tool_input" or "unavailable" or "max_uses_exceeded" or 3 more
union_member_1: array of WebSearchResultBlock { encrypted_content, page_age, title, 2 more }
tool_use_id: string
web_fetch_tool_result_block: object{ caller, content, tool_use_id, type }
caller: DirectCaller { type } or ServerToolCaller { tool_id, type } or ServerToolCaller20260120 { tool_id, type }Tool invocation directly from the model.
Tool invocation directly from the model.
direct_caller: object{ type }Tool invocation directly from the model.
Tool invocation directly from the model.
server_tool_caller: object{ tool_id, type }Tool invocation generated by a server-side tool.
Tool invocation generated by a server-side tool.
tool_id: string
server_tool_caller_20260120: object{ tool_id, type }
tool_id: string
content: WebFetchToolResultErrorBlock { error_code, type } or WebFetchBlock { content, retrieved_at, type, url }
web_fetch_tool_result_error_block: object{ error_code, type }
error_code: "invalid_tool_input" or "url_too_long" or "url_not_allowed" or 6 more
web_fetch_block: object{ content, retrieved_at, type, url }
tool_use_id: string
code_execution_tool_result_block: object{ content, tool_use_id, type }
content: CodeExecutionToolResultError { error_code, type } or CodeExecutionResultBlock { content, return_code, stderr, 2 more } or EncryptedCodeExecutionResultBlock { content, encrypted_stdout, return_code, 2 more }Code execution result with encrypted stdout for PFC + web_search results.
Code execution result with encrypted stdout for PFC + web_search results.
code_execution_tool_result_error: object{ error_code, type }
error_code: "invalid_tool_input" or "unavailable" or "too_many_requests" or "execution_time_exceeded"
code_execution_result_block: object{ content, return_code, stderr, 2 more }
content: array of CodeExecutionOutputBlock { file_id, type }
encrypted_code_execution_result_block: object{ content, encrypted_stdout, return_code, 2 more }Code execution result with encrypted stdout for PFC + web_search results.
Code execution result with encrypted stdout for PFC + web_search results.
content: array of CodeExecutionOutputBlock { file_id, type }
tool_use_id: string
bash_code_execution_tool_result_block: object{ content, tool_use_id, type }
content: BashCodeExecutionToolResultError { error_code, type } or BashCodeExecutionResultBlock { content, return_code, stderr, 2 more }
bash_code_execution_tool_result_error: object{ error_code, type }
error_code: "invalid_tool_input" or "unavailable" or "too_many_requests" or 2 more
bash_code_execution_result_block: object{ content, return_code, stderr, 2 more }
content: array of BashCodeExecutionOutputBlock { file_id, type }
tool_use_id: string
text_editor_code_execution_tool_result_block: object{ content, tool_use_id, type }
content: TextEditorCodeExecutionToolResultError { error_code, error_message, type } or TextEditorCodeExecutionViewResultBlock { content, file_type, num_lines, 3 more } or TextEditorCodeExecutionCreateResultBlock { is_file_update, type } or TextEditorCodeExecutionStrReplaceResultBlock { lines, new_lines, new_start, 3 more }
text_editor_code_execution_tool_result_error: object{ error_code, error_message, type }
error_code: "invalid_tool_input" or "unavailable" or "too_many_requests" or 2 more
text_editor_code_execution_view_result_block: object{ content, file_type, num_lines, 3 more }
file_type: "text" or "image" or "pdf"
text_editor_code_execution_create_result_block: object{ is_file_update, type }
text_editor_code_execution_str_replace_result_block: object{ lines, new_lines, new_start, 3 more }
tool_use_id: string
tool_search_tool_result_block: object{ content, tool_use_id, type }
content: ToolSearchToolResultError { error_code, error_message, type } or ToolSearchToolSearchResultBlock { tool_references, type }
tool_search_tool_result_error: object{ error_code, error_message, type }
error_code: "invalid_tool_input" or "unavailable" or "too_many_requests" or "execution_time_exceeded"
tool_search_tool_search_result_block: object{ tool_references, type }
tool_references: array of ToolReferenceBlock { tool_name, type }
tool_name: string
tool_use_id: string
container_upload_block: object{ file_id, type }Response model for a file uploaded to the container.
Response model for a file uploaded to the container.
model: "claude-fable-5-1" or "claude-mythos-5-1" or "claude-sonnet-5" or 14 more or stringThe model that will complete your prompt.
The model that will complete your prompt.
See models for additional details and options.
Frontier intelligence for ambitious tasks across coding, scientific discovery, and enterprise workflows
Our most capable model for cybersecurity and biology research, available through trusted access programs
High-performance model for coding and agents
Next generation of intelligence for the hardest knowledge work and coding problems
Most capable model for cybersecurity and biology research
Powerful intelligence for long-running agents and coding
Powerful intelligence for long-running agents and coding
Powerful intelligence for long-running agents and coding
New class of intelligence, strongest in coding and cybersecurity
Powerful intelligence for long-running agents and coding
Best combination of speed and intelligence
Fastest model with near-frontier intelligence
Fastest model with near-frontier intelligence
Powerful intelligence for long-running agents and coding
Powerful intelligence for long-running agents and coding
High-performance model for agents and coding
High-performance model for agents and coding
role: "assistant"Conversational role of the generated message.
Conversational role of the generated message.
This will always be "assistant".
stop_details: object{ category, explanation, type }Structured information about a refusal.
Structured information about a refusal.
category: "cyber" or "bio" or "frontier_llm" or 2 moreThe policy category that triggered a refusal.
The policy category that triggered a refusal.
The request could enable cyber harm, such as malware or exploit development. Benign cybersecurity work can also trigger this category.
The request could enable biological harm, such as dangerous lab methods. Beneficial life sciences work can also trigger this category.
The request could assist the development of competing AI models, which is restricted under Anthropic's commercial terms. Benign machine learning work can also trigger this category.
The request asks the model to reproduce its internal reasoning in the response text. To get reasoning in a structured form instead, use adaptive thinking.
The request could be related to an area that was determined as harmful. Benign work might sometimes trigger this category.
explanation: stringHuman-readable explanation of the refusal.
Human-readable explanation of the refusal.
This text is not guaranteed to be stable. null when no explanation is available for the category.
stop_reason: "end_turn" or "max_tokens" or "stop_sequence" or 4 moreThe reason that we stopped.
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 requestedmax_tokensor the model's maximum"stop_sequence": one of your provided customstop_sequenceswas 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.
stop_sequence: stringWhich custom stop sequence was generated, if any.
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.
type: "message"Object type.
Object type.
For Messages, this is always "message".
usage: object{ cache_creation, cache_creation_input_tokens, cache_read_input_tokens, 6 more }Billing and rate-limit 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.
cache_creation: object{ ephemeral_1h_input_tokens, ephemeral_5m_input_tokens }Breakdown of cached tokens by TTL
Breakdown of cached tokens by TTL
ephemeral_1h_input_tokens: numberThe number of input tokens used to create the 1 hour cache entry.
The number of input tokens used to create the 1 hour cache entry.
ephemeral_5m_input_tokens: numberThe number of input tokens used to create the 5 minute cache entry.
The number of input tokens used to create the 5 minute cache entry.
cache_creation_input_tokens: numberThe number of input tokens used to create the cache entry.
The number of input tokens used to create the cache entry.
cache_read_input_tokens: numberThe number of input tokens read from the cache.
The number of input tokens read from the cache.
The geographic region where inference was performed for this request.
input_tokens: numberThe number of input tokens which were used.
The number of input tokens which were used.
output_tokens: numberThe number of output tokens which were used.
The number of output tokens which were used.
output_tokens_details: object{ thinking_tokens }Breakdown of output tokens by category.
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.
thinking_tokens: numberNumber of output tokens the model generated as internal reasoning, including
the thinking-block delimiter tokens.
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. Computed by
re-tokenizing the raw reasoning text, so 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.
server_tool_use: object{ web_fetch_requests, web_search_requests }The number of server tool requests.
The number of server tool requests.
web_fetch_requests: numberThe number of web fetch tool requests.
The number of web fetch tool requests.
web_search_requests: numberThe number of web search tool requests.
The number of web search tool requests.
service_tier: "standard" or "priority" or "batch"If the request used the priority, standard, or batch tier.
If the request used the priority, standard, or batch tier.
message_batch_errored_result: object{ error, type }
error: object{ error, request_id, type }
error: InvalidRequestError { message, type } or AuthenticationError { message, type } or BillingError { message, type } or 6 more
invalid_request_error: object{ message, type }
authentication_error: object{ message, type }
billing_error: object{ message, type }
permission_error: object{ message, type }
not_found_error: object{ message, type }
rate_limit_error: object{ message, type }
gateway_timeout_error: object{ message, type }
api_error_object: object{ message, type }
overloaded_error: object{ message, type }
message_batch_canceled_result: object{ type }
message_batch_expired_result: object{ type }
ant messages:batches results \
--api-key my-anthropic-api-key \
--message-batch-id message_batch_id