Anthropic Messages API
The Messages API provides an Anthropic-compatible interface for secure LLM interactions using the Messages format.
Anthropic Messages API — client.control.messages.
Methods:
-
create–Send an Anthropic Messages API request through Sequrity.
create(
messages: list[MessageParam | dict],
model: str,
max_tokens: int,
*,
system: str | list[TextBlockParam] | list[dict] | None = None,
temperature: float | None = None,
top_p: float | None = None,
top_k: int | None = None,
tools: list[ToolParam | dict] | None = None,
tool_choice: ToolChoiceParam | dict | None = None,
thinking: ThinkingConfigParam | dict | None = None,
reasoning_effort: ReasoningEffort | None = None,
stop_sequences: list[str] | None = None,
stream: Literal[True],
output_config: OutputConfigParam | dict | None = None,
metadata: MetadataParam | dict | None = None,
service_tier: Literal["auto", "standard_only"] | None = None,
timeout: float | None = None,
provider: LlmServiceProvider | LlmServiceProviderStr | None | _NotGiven = NOT_GIVEN,
llm_api_key: str | None | _NotGiven = NOT_GIVEN,
features: FeaturesHeader | None | _NotGiven = NOT_GIVEN,
security_policy: SecurityPolicyHeader | None | _NotGiven = NOT_GIVEN,
fine_grained_config: FineGrainedConfigHeader | None | _NotGiven = NOT_GIVEN,
endpoint_type: str | _NotGiven = NOT_GIVEN,
session_id: str | None | _NotGiven = NOT_GIVEN,
feature_overrides: dict[str, Any] | None = None,
policy_overrides: dict[str, Any] | None = None,
config_overrides: dict[str, Any] | None = None,
custom_headers: dict[str, str] | None = None,
) -> SyncStream[AnthropicStreamEvent]
create(
messages: list[MessageParam | dict],
model: str,
max_tokens: int,
*,
system: str | list[TextBlockParam] | list[dict] | None = None,
temperature: float | None = None,
top_p: float | None = None,
top_k: int | None = None,
tools: list[ToolParam | dict] | None = None,
tool_choice: ToolChoiceParam | dict | None = None,
thinking: ThinkingConfigParam | dict | None = None,
reasoning_effort: ReasoningEffort | None = None,
stop_sequences: list[str] | None = None,
stream: Literal[False] | None = None,
output_config: OutputConfigParam | dict | None = None,
metadata: MetadataParam | dict | None = None,
service_tier: Literal["auto", "standard_only"] | None = None,
timeout: float | None = None,
provider: LlmServiceProvider | LlmServiceProviderStr | None | _NotGiven = NOT_GIVEN,
llm_api_key: str | None | _NotGiven = NOT_GIVEN,
features: FeaturesHeader | None | _NotGiven = NOT_GIVEN,
security_policy: SecurityPolicyHeader | None | _NotGiven = NOT_GIVEN,
fine_grained_config: FineGrainedConfigHeader | None | _NotGiven = NOT_GIVEN,
endpoint_type: str | _NotGiven = NOT_GIVEN,
session_id: str | None | _NotGiven = NOT_GIVEN,
feature_overrides: dict[str, Any] | None = None,
policy_overrides: dict[str, Any] | None = None,
config_overrides: dict[str, Any] | None = None,
custom_headers: dict[str, str] | None = None,
) -> AnthropicMessageResponse
create(
messages: list[MessageParam | dict],
model: str,
max_tokens: int,
*,
system: str | list[TextBlockParam] | list[dict] | None = None,
temperature: float | None = None,
top_p: float | None = None,
top_k: int | None = None,
tools: list[ToolParam | dict] | None = None,
tool_choice: ToolChoiceParam | dict | None = None,
thinking: ThinkingConfigParam | dict | None = None,
reasoning_effort: ReasoningEffort | None = None,
stop_sequences: list[str] | None = None,
stream: bool | None = None,
output_config: OutputConfigParam | dict | None = None,
metadata: MetadataParam | dict | None = None,
service_tier: Literal["auto", "standard_only"] | None = None,
timeout: float | None = None,
provider: LlmServiceProvider | LlmServiceProviderStr | None | _NotGiven = NOT_GIVEN,
llm_api_key: str | None | _NotGiven = NOT_GIVEN,
features: FeaturesHeader | None | _NotGiven = NOT_GIVEN,
security_policy: SecurityPolicyHeader | None | _NotGiven = NOT_GIVEN,
fine_grained_config: FineGrainedConfigHeader | None | _NotGiven = NOT_GIVEN,
endpoint_type: str | _NotGiven = NOT_GIVEN,
session_id: str | None | _NotGiven = NOT_GIVEN,
feature_overrides: dict[str, Any] | None = None,
policy_overrides: dict[str, Any] | None = None,
config_overrides: dict[str, Any] | None = None,
custom_headers: dict[str, str] | None = None,
) -> AnthropicMessageResponse | SyncStream[AnthropicStreamEvent]
Send an Anthropic Messages API request through Sequrity.
Parameters:
-
(messageslist[MessageParam | dict]) –List of messages with alternating user and assistant turns.
-
(modelstr) –The model identifier (e.g.,
"claude-4-sonnet"). -
(max_tokensint) –Maximum number of tokens to generate before stopping.
-
(systemstr | list[TextBlockParam] | list[dict] | None, default:None) –System prompt as a string or list of text blocks.
-
(temperaturefloat | None, default:None) –Sampling temperature (0.0 to 1.0).
-
(top_pfloat | None, default:None) –Nucleus sampling probability threshold.
-
(top_kint | None, default:None) –Only sample from the top K options for each token.
-
(toolslist[ToolParam | dict] | None, default:None) –List of tool definitions available to the model.
-
(tool_choiceToolChoiceParam | dict | None, default:None) –How the model should use the provided tools.
-
(thinkingThinkingConfigParam | dict | None, default:None) –Configuration for extended thinking.
-
(reasoning_effortReasoningEffort | None, default:None) –Reasoning effort level for supported models.
-
(stop_sequenceslist[str] | None, default:None) –Custom text sequences that cause the model to stop.
-
(streambool | None, default:None) –Whether to stream the response. When
True, returns a :class:SyncStreamof :class:AnthropicStreamEventobjects. -
(output_configOutputConfigParam | dict | None, default:None) –Output format configuration.
-
(metadataMetadataParam | dict | None, default:None) –Request metadata.
-
(service_tierLiteral['auto', 'standard_only'] | None, default:None) –Priority or standard capacity selection.
-
(timeoutfloat | None, default:None) –Maximum request duration in seconds.
-
(providerLlmServiceProvider | LlmServiceProviderStr | None | _NotGiven, default:NOT_GIVEN) –LLM service provider override.
-
(llm_api_keystr | None | _NotGiven, default:NOT_GIVEN) –LLM provider API key override.
-
(featuresFeaturesHeader | None | _NotGiven, default:NOT_GIVEN) –Security features override.
-
(security_policySecurityPolicyHeader | None | _NotGiven, default:NOT_GIVEN) –Security policy override.
-
(fine_grained_configFineGrainedConfigHeader | None | _NotGiven, default:NOT_GIVEN) –Fine-grained config override.
-
(endpoint_typestr | _NotGiven, default:NOT_GIVEN) –Endpoint type override.
-
(session_idstr | None | _NotGiven, default:NOT_GIVEN) –Explicit session ID override.
-
(feature_overridesdict[str, Any] | None, default:None) –Dict to deep-merge into the serialized
X-Featuresheader JSON. Allows adding or overriding fields without loosening Pydantic validation on :class:FeaturesHeader. -
(policy_overridesdict[str, Any] | None, default:None) –Dict to deep-merge into the serialized
X-Policyheader JSON. -
(config_overridesdict[str, Any] | None, default:None) –Dict to deep-merge into the serialized
X-Configheader JSON.
Returns:
-
AnthropicMessageResponse | SyncStream[AnthropicStreamEvent]–AnthropicMessageResponsewhenstreamisFalse/None, -
AnthropicMessageResponse | SyncStream[AnthropicStreamEvent]–or
SyncStream[AnthropicStreamEvent]whenstreamisTrue.
Request Types
pydantic-model
Anthropic Messages API request.
This is a Pydantic model representation of the Anthropic Messages API request, used for sending messages through Sequrity's Control API with the Messages format.
Show JSON schema:
{
"$defs": {
"Base64DocumentSourceParam": {
"description": "Base64-encoded document source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of document source.",
"title": "Type",
"type": "string"
},
"media_type": {
"const": "application/pdf",
"description": "The media type of the document.",
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded document data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64DocumentSourceParam",
"type": "object"
},
"Base64ImageSourceParam": {
"description": "Base64-encoded image source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"media_type": {
"description": "The media type of the image.",
"enum": [
"image/jpeg",
"image/png",
"image/gif",
"image/webp"
],
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded image data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64ImageSourceParam",
"type": "object"
},
"DocumentBlockParam": {
"description": "Document content block (PDF).",
"properties": {
"source": {
"$ref": "#/$defs/Base64DocumentSourceParam",
"description": "The source of the document."
},
"type": {
"const": "document",
"default": "document",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "DocumentBlockParam",
"type": "object"
},
"ImageBlockParam": {
"description": "Image content block.",
"properties": {
"source": {
"description": "The source of the image.",
"discriminator": {
"mapping": {
"base64": "#/$defs/Base64ImageSourceParam",
"url": "#/$defs/URLImageSourceParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/Base64ImageSourceParam"
},
{
"$ref": "#/$defs/URLImageSourceParam"
}
],
"title": "Source"
},
"type": {
"const": "image",
"default": "image",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "ImageBlockParam",
"type": "object"
},
"JSONOutputFormatParam": {
"additionalProperties": false,
"description": "JSON output format with schema.",
"properties": {
"type": {
"const": "json_schema",
"default": "json_schema",
"description": "The type of the output format.",
"title": "Type",
"type": "string"
},
"schema": {
"additionalProperties": true,
"description": "The JSON schema of the format.",
"title": "Schema",
"type": "object"
}
},
"required": [
"schema"
],
"title": "JSONOutputFormatParam",
"type": "object"
},
"MessageParam": {
"description": "A message in the conversation.",
"properties": {
"role": {
"description": "The role of the message author: 'user' or 'assistant'.",
"enum": [
"user",
"assistant"
],
"title": "Role",
"type": "string"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"discriminator": {
"mapping": {
"document": "#/$defs/DocumentBlockParam",
"image": "#/$defs/ImageBlockParam",
"text": "#/$defs/TextBlockParam",
"thinking": "#/$defs/ThinkingBlockParam",
"tool_result": "#/$defs/ToolResultBlockParam",
"tool_use": "#/$defs/ToolUseBlockParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/TextBlockParam"
},
{
"$ref": "#/$defs/ImageBlockParam"
},
{
"$ref": "#/$defs/DocumentBlockParam"
},
{
"$ref": "#/$defs/ToolUseBlockParam"
},
{
"$ref": "#/$defs/ToolResultBlockParam"
},
{
"$ref": "#/$defs/ThinkingBlockParam"
}
]
},
"type": "array"
}
],
"description": "The contents of the message. Can be a string or an array of content blocks.",
"title": "Content"
}
},
"required": [
"role",
"content"
],
"title": "MessageParam",
"type": "object"
},
"MetadataParam": {
"description": "Request metadata.",
"properties": {
"user_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "An external identifier for the user associated with the request. Should be a uuid, hash, or other opaque identifier.",
"title": "User Id"
}
},
"title": "MetadataParam",
"type": "object"
},
"OutputConfigParam": {
"description": "Output configuration for structured responses.",
"properties": {
"format": {
"anyOf": [
{
"$ref": "#/$defs/JSONOutputFormatParam"
},
{
"type": "null"
}
],
"default": null,
"description": "A schema to specify Claude's output format in responses."
}
},
"title": "OutputConfigParam",
"type": "object"
},
"TextBlockParam": {
"description": "Text content block.",
"properties": {
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "text",
"default": "text",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"citations": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"text"
],
"title": "TextBlockParam",
"type": "object"
},
"ThinkingBlockParam": {
"description": "Thinking content block for extended thinking.",
"properties": {
"text": {
"description": "The thinking content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "thinking",
"default": "thinking",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"text"
],
"title": "ThinkingBlockParam",
"type": "object"
},
"ThinkingConfigAdaptiveParam": {
"description": "Adaptive thinking configuration.\n\nClaude automatically decides whether to use extended thinking based on\nthe complexity of the request.\nSee https://docs.anthropic.com/en/api/messages",
"properties": {
"type": {
"const": "adaptive",
"default": "adaptive",
"description": "Thinking is adaptive.",
"title": "Type",
"type": "string"
}
},
"title": "ThinkingConfigAdaptiveParam",
"type": "object"
},
"ThinkingConfigDisabledParam": {
"description": "Extended thinking disabled configuration.",
"properties": {
"type": {
"const": "disabled",
"default": "disabled",
"description": "Thinking is disabled.",
"title": "Type",
"type": "string"
}
},
"title": "ThinkingConfigDisabledParam",
"type": "object"
},
"ThinkingConfigEnabledParam": {
"description": "Extended thinking enabled configuration.",
"properties": {
"type": {
"const": "enabled",
"default": "enabled",
"description": "Thinking is enabled.",
"title": "Type",
"type": "string"
},
"budget_tokens": {
"description": "The maximum number of tokens to use for thinking. Minimum is 1024.",
"title": "Budget Tokens",
"type": "integer"
}
},
"required": [
"budget_tokens"
],
"title": "ThinkingConfigEnabledParam",
"type": "object"
},
"ToolChoiceAnyParam": {
"description": "Any tool choice \u2014 model must use at least one tool.",
"properties": {
"type": {
"const": "any",
"default": "any",
"description": "The model will use any available tools.",
"title": "Type",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output exactly one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"title": "ToolChoiceAnyParam",
"type": "object"
},
"ToolChoiceAutoParam": {
"description": "Auto tool choice \u2014 model decides whether to use tools.",
"properties": {
"type": {
"const": "auto",
"default": "auto",
"description": "The model will automatically decide whether to use tools.",
"title": "Type",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output at most one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"title": "ToolChoiceAutoParam",
"type": "object"
},
"ToolChoiceNoneParam": {
"description": "None tool choice \u2014 model will not use tools.",
"properties": {
"type": {
"const": "none",
"default": "none",
"description": "The model will not use any tools.",
"title": "Type",
"type": "string"
}
},
"title": "ToolChoiceNoneParam",
"type": "object"
},
"ToolChoiceToolParam": {
"description": "Specific tool choice \u2014 model must use the named tool.",
"properties": {
"type": {
"const": "tool",
"default": "tool",
"description": "The model will use the specified tool.",
"title": "Type",
"type": "string"
},
"name": {
"description": "The name of the tool to use.",
"title": "Name",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output exactly one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"required": [
"name"
],
"title": "ToolChoiceToolParam",
"type": "object"
},
"ToolInputSchema": {
"additionalProperties": true,
"description": "JSON Schema for tool input parameters.",
"properties": {
"type": {
"const": "object",
"default": "object",
"description": "The type of the input schema. Always 'object'.",
"title": "Type",
"type": "string"
},
"properties": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "The properties of the input schema as a JSON Schema object.",
"title": "Properties"
},
"required": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The required properties of the input schema.",
"title": "Required"
}
},
"title": "ToolInputSchema",
"type": "object"
},
"ToolParam": {
"description": "Tool definition for the Anthropic Messages API.",
"properties": {
"name": {
"description": "Name of the tool. This is how the tool will be called by the model.",
"title": "Name",
"type": "string"
},
"input_schema": {
"anyOf": [
{
"$ref": "#/$defs/ToolInputSchema"
},
{
"additionalProperties": true,
"type": "object"
}
],
"description": "JSON schema for this tool's input. Defines the shape of the input that the model will produce.",
"title": "Input Schema"
},
"description": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Description of what this tool does. Tool descriptions should be as detailed as possible.",
"title": "Description"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"strict": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "When true, guarantees schema validation on tool names and inputs.",
"title": "Strict"
},
"type": {
"anyOf": [
{
"const": "custom",
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "The type of the tool.",
"title": "Type"
}
},
"required": [
"name",
"input_schema"
],
"title": "ToolParam",
"type": "object"
},
"ToolResultBlockParam": {
"description": "Tool result content block in user messages.",
"properties": {
"tool_use_id": {
"description": "The id of the tool use request this is a result for. Must match an id from a previous tool_use block.",
"title": "Tool Use Id",
"type": "string"
},
"type": {
"const": "tool_result",
"default": "tool_result",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"anyOf": [
{
"$ref": "#/$defs/TextBlockParam"
},
{
"$ref": "#/$defs/ImageBlockParam"
},
{
"$ref": "#/$defs/DocumentBlockParam"
}
]
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The result of the tool call.",
"title": "Content"
},
"is_error": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Set to true if the tool execution resulted in an error.",
"title": "Is Error"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"tool_use_id"
],
"title": "ToolResultBlockParam",
"type": "object"
},
"ToolUseBlockParam": {
"description": "Tool use content block in assistant messages.",
"properties": {
"id": {
"description": "A unique identifier for this particular tool use block.",
"title": "Id",
"type": "string"
},
"input": {
"additionalProperties": true,
"description": "An object containing the input being passed to the tool.",
"title": "Input",
"type": "object"
},
"name": {
"description": "The name of the tool being used.",
"title": "Name",
"type": "string"
},
"type": {
"const": "tool_use",
"default": "tool_use",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"id",
"input",
"name"
],
"title": "ToolUseBlockParam",
"type": "object"
},
"URLImageSourceParam": {
"description": "URL image source.",
"properties": {
"type": {
"const": "url",
"default": "url",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"url": {
"description": "The URL of the image.",
"title": "Url",
"type": "string"
}
},
"required": [
"url"
],
"title": "URLImageSourceParam",
"type": "object"
}
},
"description": "Anthropic Messages API request.\n\nThis is a Pydantic model representation of the Anthropic Messages API request,\nused for sending messages through Sequrity's Control API with the Messages format.",
"properties": {
"max_tokens": {
"description": "The maximum number of tokens to generate before stopping.",
"title": "Max Tokens",
"type": "integer"
},
"messages": {
"description": "Input messages with alternating user and assistant conversational turns.",
"items": {
"$ref": "#/$defs/MessageParam"
},
"title": "Messages",
"type": "array"
},
"model": {
"description": "The model that will complete your prompt (e.g., 'claude-3-opus', 'claude-4-sonnet').",
"title": "Model",
"type": "string"
},
"metadata": {
"anyOf": [
{
"$ref": "#/$defs/MetadataParam"
},
{
"type": "null"
}
],
"default": null,
"description": "An object describing metadata about the request."
},
"reasoning_effort": {
"anyOf": [
{
"enum": [
"none",
"minimal",
"low",
"medium",
"high",
"xhigh"
],
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Constrains effort on reasoning for reasoning models. Supported values are 'none', 'minimal', 'low', 'medium', 'high', and 'xhigh'.",
"title": "Reasoning Effort"
},
"output_config": {
"anyOf": [
{
"$ref": "#/$defs/OutputConfigParam"
},
{
"type": "null"
}
],
"default": null,
"description": "Configuration options for the model's output, such as the output format."
},
"service_tier": {
"anyOf": [
{
"enum": [
"auto",
"standard_only"
],
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Determines whether to use priority capacity or standard capacity for this request.",
"title": "Service Tier"
},
"stop_sequences": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Custom text sequences that will cause the model to stop generating.",
"title": "Stop Sequences"
},
"stream": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to incrementally stream the response using server-sent events.",
"title": "Stream"
},
"system": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"$ref": "#/$defs/TextBlockParam"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "System prompt providing context and instructions to the model.",
"title": "System"
},
"temperature": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"default": null,
"description": "Amount of randomness injected into the response. Ranges from 0.0 to 1.0.",
"title": "Temperature"
},
"thinking": {
"anyOf": [
{
"discriminator": {
"mapping": {
"adaptive": "#/$defs/ThinkingConfigAdaptiveParam",
"disabled": "#/$defs/ThinkingConfigDisabledParam",
"enabled": "#/$defs/ThinkingConfigEnabledParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/ThinkingConfigEnabledParam"
},
{
"$ref": "#/$defs/ThinkingConfigDisabledParam"
},
{
"$ref": "#/$defs/ThinkingConfigAdaptiveParam"
}
]
},
{
"type": "null"
}
],
"default": null,
"description": "Configuration for enabling extended thinking. When enabled, responses include thinking content blocks.",
"title": "Thinking"
},
"tool_choice": {
"anyOf": [
{
"discriminator": {
"mapping": {
"any": "#/$defs/ToolChoiceAnyParam",
"auto": "#/$defs/ToolChoiceAutoParam",
"none": "#/$defs/ToolChoiceNoneParam",
"tool": "#/$defs/ToolChoiceToolParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/ToolChoiceAutoParam"
},
{
"$ref": "#/$defs/ToolChoiceAnyParam"
},
{
"$ref": "#/$defs/ToolChoiceToolParam"
},
{
"$ref": "#/$defs/ToolChoiceNoneParam"
}
]
},
{
"type": "null"
}
],
"default": null,
"description": "How the model should use the provided tools.",
"title": "Tool Choice"
},
"tools": {
"anyOf": [
{
"items": {
"anyOf": [
{
"$ref": "#/$defs/ToolParam"
},
{
"additionalProperties": true,
"type": "object"
}
]
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Definitions of tools that the model may use.",
"title": "Tools"
},
"top_k": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"description": "Only sample from the top K options for each subsequent token.",
"title": "Top K"
},
"top_p": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"default": null,
"description": "Use nucleus sampling with the specified probability threshold.",
"title": "Top P"
},
"timeout": {
"anyOf": [
{
"type": "number"
},
{
"type": "null"
}
],
"default": null,
"description": "The maximum amount of time, in seconds, that the request should take.",
"title": "Timeout"
}
},
"required": [
"max_tokens",
"messages",
"model"
],
"title": "AnthropicMessageRequest",
"type": "object"
}
Fields:
-
max_tokens(int) -
messages(list[MessageParam]) -
model(str) -
metadata(MetadataParam | None) -
reasoning_effort(ReasoningEffort | None) -
output_config(OutputConfigParam | None) -
service_tier(Literal['auto', 'standard_only'] | None) -
stop_sequences(list[str] | None) -
stream(bool | None) -
system(str | list[TextBlockParam] | None) -
temperature(float | None) -
thinking(ThinkingConfigParam | None) -
tool_choice(ToolChoiceParam | None) -
tools(list[ToolUnionParam] | None) -
top_k(int | None) -
top_p(float | None) -
timeout(float | None)
pydantic-field
messages: list[MessageParam]
Input messages with alternating user and assistant conversational turns.
pydantic-field
model: str
The model that will complete your prompt (e.g., 'claude-3-opus', 'claude-4-sonnet').
pydantic-field
An object describing metadata about the request.
pydantic-field
Constrains effort on reasoning for reasoning models. Supported values are 'none', 'minimal', 'low', 'medium', 'high', and 'xhigh'.
pydantic-field
output_config: OutputConfigParam | None = None
Configuration options for the model's output, such as the output format.
pydantic-field
service_tier: Literal['auto', 'standard_only'] | None = None
Determines whether to use priority capacity or standard capacity for this request.
pydantic-field
Custom text sequences that will cause the model to stop generating.
pydantic-field
stream: bool | None = None
Whether to incrementally stream the response using server-sent events.
pydantic-field
system: str | list[TextBlockParam] | None = None
System prompt providing context and instructions to the model.
pydantic-field
temperature: float | None = None
Amount of randomness injected into the response. Ranges from 0.0 to 1.0.
pydantic-field
Configuration for enabling extended thinking. When enabled, responses include thinking content blocks.
pydantic-field
How the model should use the provided tools.
pydantic-field
tools: list[ToolUnionParam] | None = None
Definitions of tools that the model may use.
pydantic-field
top_k: int | None = None
Only sample from the top K options for each subsequent token.
pydantic-field
top_p: float | None = None
Use nucleus sampling with the specified probability threshold.
pydantic-field
timeout: float | None = None
The maximum amount of time, in seconds, that the request should take.
pydantic-model
A message in the conversation.
Show JSON schema:
{
"$defs": {
"Base64DocumentSourceParam": {
"description": "Base64-encoded document source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of document source.",
"title": "Type",
"type": "string"
},
"media_type": {
"const": "application/pdf",
"description": "The media type of the document.",
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded document data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64DocumentSourceParam",
"type": "object"
},
"Base64ImageSourceParam": {
"description": "Base64-encoded image source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"media_type": {
"description": "The media type of the image.",
"enum": [
"image/jpeg",
"image/png",
"image/gif",
"image/webp"
],
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded image data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64ImageSourceParam",
"type": "object"
},
"DocumentBlockParam": {
"description": "Document content block (PDF).",
"properties": {
"source": {
"$ref": "#/$defs/Base64DocumentSourceParam",
"description": "The source of the document."
},
"type": {
"const": "document",
"default": "document",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "DocumentBlockParam",
"type": "object"
},
"ImageBlockParam": {
"description": "Image content block.",
"properties": {
"source": {
"description": "The source of the image.",
"discriminator": {
"mapping": {
"base64": "#/$defs/Base64ImageSourceParam",
"url": "#/$defs/URLImageSourceParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/Base64ImageSourceParam"
},
{
"$ref": "#/$defs/URLImageSourceParam"
}
],
"title": "Source"
},
"type": {
"const": "image",
"default": "image",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "ImageBlockParam",
"type": "object"
},
"TextBlockParam": {
"description": "Text content block.",
"properties": {
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "text",
"default": "text",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"citations": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"text"
],
"title": "TextBlockParam",
"type": "object"
},
"ThinkingBlockParam": {
"description": "Thinking content block for extended thinking.",
"properties": {
"text": {
"description": "The thinking content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "thinking",
"default": "thinking",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"text"
],
"title": "ThinkingBlockParam",
"type": "object"
},
"ToolResultBlockParam": {
"description": "Tool result content block in user messages.",
"properties": {
"tool_use_id": {
"description": "The id of the tool use request this is a result for. Must match an id from a previous tool_use block.",
"title": "Tool Use Id",
"type": "string"
},
"type": {
"const": "tool_result",
"default": "tool_result",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"anyOf": [
{
"$ref": "#/$defs/TextBlockParam"
},
{
"$ref": "#/$defs/ImageBlockParam"
},
{
"$ref": "#/$defs/DocumentBlockParam"
}
]
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The result of the tool call.",
"title": "Content"
},
"is_error": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Set to true if the tool execution resulted in an error.",
"title": "Is Error"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"tool_use_id"
],
"title": "ToolResultBlockParam",
"type": "object"
},
"ToolUseBlockParam": {
"description": "Tool use content block in assistant messages.",
"properties": {
"id": {
"description": "A unique identifier for this particular tool use block.",
"title": "Id",
"type": "string"
},
"input": {
"additionalProperties": true,
"description": "An object containing the input being passed to the tool.",
"title": "Input",
"type": "object"
},
"name": {
"description": "The name of the tool being used.",
"title": "Name",
"type": "string"
},
"type": {
"const": "tool_use",
"default": "tool_use",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"id",
"input",
"name"
],
"title": "ToolUseBlockParam",
"type": "object"
},
"URLImageSourceParam": {
"description": "URL image source.",
"properties": {
"type": {
"const": "url",
"default": "url",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"url": {
"description": "The URL of the image.",
"title": "Url",
"type": "string"
}
},
"required": [
"url"
],
"title": "URLImageSourceParam",
"type": "object"
}
},
"description": "A message in the conversation.",
"properties": {
"role": {
"description": "The role of the message author: 'user' or 'assistant'.",
"enum": [
"user",
"assistant"
],
"title": "Role",
"type": "string"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"discriminator": {
"mapping": {
"document": "#/$defs/DocumentBlockParam",
"image": "#/$defs/ImageBlockParam",
"text": "#/$defs/TextBlockParam",
"thinking": "#/$defs/ThinkingBlockParam",
"tool_result": "#/$defs/ToolResultBlockParam",
"tool_use": "#/$defs/ToolUseBlockParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/TextBlockParam"
},
{
"$ref": "#/$defs/ImageBlockParam"
},
{
"$ref": "#/$defs/DocumentBlockParam"
},
{
"$ref": "#/$defs/ToolUseBlockParam"
},
{
"$ref": "#/$defs/ToolResultBlockParam"
},
{
"$ref": "#/$defs/ThinkingBlockParam"
}
]
},
"type": "array"
}
],
"description": "The contents of the message. Can be a string or an array of content blocks.",
"title": "Content"
}
},
"required": [
"role",
"content"
],
"title": "MessageParam",
"type": "object"
}
Config:
extra:ignore
Fields:
Content Blocks
pydantic-model
Text content block.
Show JSON schema:
{
"description": "Text content block.",
"properties": {
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "text",
"default": "text",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"citations": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"text"
],
"title": "TextBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
text(str) -
type(Literal['text']) -
cache_control(dict[str, Any] | None) -
citations(list[dict[str, Any]] | None)
pydantic-field
Create a cache control breakpoint at this content block.
pydantic-model
Image content block.
Show JSON schema:
{
"$defs": {
"Base64ImageSourceParam": {
"description": "Base64-encoded image source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"media_type": {
"description": "The media type of the image.",
"enum": [
"image/jpeg",
"image/png",
"image/gif",
"image/webp"
],
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded image data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64ImageSourceParam",
"type": "object"
},
"URLImageSourceParam": {
"description": "URL image source.",
"properties": {
"type": {
"const": "url",
"default": "url",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"url": {
"description": "The URL of the image.",
"title": "Url",
"type": "string"
}
},
"required": [
"url"
],
"title": "URLImageSourceParam",
"type": "object"
}
},
"description": "Image content block.",
"properties": {
"source": {
"description": "The source of the image.",
"discriminator": {
"mapping": {
"base64": "#/$defs/Base64ImageSourceParam",
"url": "#/$defs/URLImageSourceParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/Base64ImageSourceParam"
},
{
"$ref": "#/$defs/URLImageSourceParam"
}
],
"title": "Source"
},
"type": {
"const": "image",
"default": "image",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "ImageBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-model
Document content block (PDF).
Show JSON schema:
{
"$defs": {
"Base64DocumentSourceParam": {
"description": "Base64-encoded document source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of document source.",
"title": "Type",
"type": "string"
},
"media_type": {
"const": "application/pdf",
"description": "The media type of the document.",
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded document data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64DocumentSourceParam",
"type": "object"
}
},
"description": "Document content block (PDF).",
"properties": {
"source": {
"$ref": "#/$defs/Base64DocumentSourceParam",
"description": "The source of the document."
},
"type": {
"const": "document",
"default": "document",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "DocumentBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-model
Tool use content block in assistant messages.
Show JSON schema:
{
"description": "Tool use content block in assistant messages.",
"properties": {
"id": {
"description": "A unique identifier for this particular tool use block.",
"title": "Id",
"type": "string"
},
"input": {
"additionalProperties": true,
"description": "An object containing the input being passed to the tool.",
"title": "Input",
"type": "object"
},
"name": {
"description": "The name of the tool being used.",
"title": "Name",
"type": "string"
},
"type": {
"const": "tool_use",
"default": "tool_use",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"id",
"input",
"name"
],
"title": "ToolUseBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
id(str) -
input(dict[str, Any]) -
name(str) -
type(Literal['tool_use']) -
cache_control(dict[str, Any] | None)
pydantic-model
Tool result content block in user messages.
Show JSON schema:
{
"$defs": {
"Base64DocumentSourceParam": {
"description": "Base64-encoded document source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of document source.",
"title": "Type",
"type": "string"
},
"media_type": {
"const": "application/pdf",
"description": "The media type of the document.",
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded document data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64DocumentSourceParam",
"type": "object"
},
"Base64ImageSourceParam": {
"description": "Base64-encoded image source.",
"properties": {
"type": {
"const": "base64",
"default": "base64",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"media_type": {
"description": "The media type of the image.",
"enum": [
"image/jpeg",
"image/png",
"image/gif",
"image/webp"
],
"title": "Media Type",
"type": "string"
},
"data": {
"description": "The base64-encoded image data.",
"title": "Data",
"type": "string"
}
},
"required": [
"media_type",
"data"
],
"title": "Base64ImageSourceParam",
"type": "object"
},
"DocumentBlockParam": {
"description": "Document content block (PDF).",
"properties": {
"source": {
"$ref": "#/$defs/Base64DocumentSourceParam",
"description": "The source of the document."
},
"type": {
"const": "document",
"default": "document",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "DocumentBlockParam",
"type": "object"
},
"ImageBlockParam": {
"description": "Image content block.",
"properties": {
"source": {
"description": "The source of the image.",
"discriminator": {
"mapping": {
"base64": "#/$defs/Base64ImageSourceParam",
"url": "#/$defs/URLImageSourceParam"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/Base64ImageSourceParam"
},
{
"$ref": "#/$defs/URLImageSourceParam"
}
],
"title": "Source"
},
"type": {
"const": "image",
"default": "image",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"source"
],
"title": "ImageBlockParam",
"type": "object"
},
"TextBlockParam": {
"description": "Text content block.",
"properties": {
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "text",
"default": "text",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"citations": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"text"
],
"title": "TextBlockParam",
"type": "object"
},
"URLImageSourceParam": {
"description": "URL image source.",
"properties": {
"type": {
"const": "url",
"default": "url",
"description": "The type of image source.",
"title": "Type",
"type": "string"
},
"url": {
"description": "The URL of the image.",
"title": "Url",
"type": "string"
}
},
"required": [
"url"
],
"title": "URLImageSourceParam",
"type": "object"
}
},
"description": "Tool result content block in user messages.",
"properties": {
"tool_use_id": {
"description": "The id of the tool use request this is a result for. Must match an id from a previous tool_use block.",
"title": "Tool Use Id",
"type": "string"
},
"type": {
"const": "tool_result",
"default": "tool_result",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"anyOf": [
{
"$ref": "#/$defs/TextBlockParam"
},
{
"$ref": "#/$defs/ImageBlockParam"
},
{
"$ref": "#/$defs/DocumentBlockParam"
}
]
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The result of the tool call.",
"title": "Content"
},
"is_error": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Set to true if the tool execution resulted in an error.",
"title": "Is Error"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"tool_use_id"
],
"title": "ToolResultBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
tool_use_id(str) -
type(Literal['tool_result']) -
content(str | list[TextBlockParam | ImageBlockParam | DocumentBlockParam] | None) -
is_error(bool | None) -
cache_control(dict[str, Any] | None)
pydantic-field
tool_use_id: str
The id of the tool use request this is a result for. Must match an id from a previous tool_use block.
pydantic-field
content: str | list[TextBlockParam | ImageBlockParam | DocumentBlockParam] | None = None
The result of the tool call.
pydantic-field
is_error: bool | None = None
Set to true if the tool execution resulted in an error.
pydantic-model
Thinking content block for extended thinking.
Show JSON schema:
{
"description": "Thinking content block for extended thinking.",
"properties": {
"text": {
"description": "The thinking content.",
"title": "Text",
"type": "string"
},
"type": {
"const": "thinking",
"default": "thinking",
"description": "The type of the content part.",
"title": "Type",
"type": "string"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
}
},
"required": [
"text"
],
"title": "ThinkingBlockParam",
"type": "object"
}
Config:
extra:ignore
Fields:
Tool Types
pydantic-model
Tool definition for the Anthropic Messages API.
Show JSON schema:
{
"$defs": {
"ToolInputSchema": {
"additionalProperties": true,
"description": "JSON Schema for tool input parameters.",
"properties": {
"type": {
"const": "object",
"default": "object",
"description": "The type of the input schema. Always 'object'.",
"title": "Type",
"type": "string"
},
"properties": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "The properties of the input schema as a JSON Schema object.",
"title": "Properties"
},
"required": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The required properties of the input schema.",
"title": "Required"
}
},
"title": "ToolInputSchema",
"type": "object"
}
},
"description": "Tool definition for the Anthropic Messages API.",
"properties": {
"name": {
"description": "Name of the tool. This is how the tool will be called by the model.",
"title": "Name",
"type": "string"
},
"input_schema": {
"anyOf": [
{
"$ref": "#/$defs/ToolInputSchema"
},
{
"additionalProperties": true,
"type": "object"
}
],
"description": "JSON schema for this tool's input. Defines the shape of the input that the model will produce.",
"title": "Input Schema"
},
"description": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Description of what this tool does. Tool descriptions should be as detailed as possible.",
"title": "Description"
},
"cache_control": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Create a cache control breakpoint at this content block.",
"title": "Cache Control"
},
"strict": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "When true, guarantees schema validation on tool names and inputs.",
"title": "Strict"
},
"type": {
"anyOf": [
{
"const": "custom",
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "The type of the tool.",
"title": "Type"
}
},
"required": [
"name",
"input_schema"
],
"title": "ToolParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
name(str) -
input_schema(ToolInputSchema | dict[str, Any]) -
description(str | None) -
cache_control(dict[str, Any] | None) -
strict(bool | None) -
type(Literal['custom'] | None)
pydantic-field
input_schema: ToolInputSchema | dict[str, Any]
JSON schema for this tool's input. Defines the shape of the input that the model will produce.
pydantic-field
description: str | None = None
Description of what this tool does. Tool descriptions should be as detailed as possible.
pydantic-field
Create a cache control breakpoint at this content block.
pydantic-field
strict: bool | None = None
When true, guarantees schema validation on tool names and inputs.
pydantic-model
JSON Schema for tool input parameters.
Show JSON schema:
{
"additionalProperties": true,
"description": "JSON Schema for tool input parameters.",
"properties": {
"type": {
"const": "object",
"default": "object",
"description": "The type of the input schema. Always 'object'.",
"title": "Type",
"type": "string"
},
"properties": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "The properties of the input schema as a JSON Schema object.",
"title": "Properties"
},
"required": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "The required properties of the input schema.",
"title": "Required"
}
},
"title": "ToolInputSchema",
"type": "object"
}
Config:
extra:allow
Fields:
pydantic-field
type: Literal['object'] = 'object'
The type of the input schema. Always 'object'.
pydantic-field
The properties of the input schema as a JSON Schema object.
pydantic-model
Auto tool choice — model decides whether to use tools.
Show JSON schema:
{
"description": "Auto tool choice \u2014 model decides whether to use tools.",
"properties": {
"type": {
"const": "auto",
"default": "auto",
"description": "The model will automatically decide whether to use tools.",
"title": "Type",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output at most one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"title": "ToolChoiceAutoParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
type(Literal['auto']) -
disable_parallel_tool_use(bool | None)
pydantic-model
Any tool choice — model must use at least one tool.
Show JSON schema:
{
"description": "Any tool choice \u2014 model must use at least one tool.",
"properties": {
"type": {
"const": "any",
"default": "any",
"description": "The model will use any available tools.",
"title": "Type",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output exactly one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"title": "ToolChoiceAnyParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
type(Literal['any']) -
disable_parallel_tool_use(bool | None)
pydantic-model
Specific tool choice — model must use the named tool.
Show JSON schema:
{
"description": "Specific tool choice \u2014 model must use the named tool.",
"properties": {
"type": {
"const": "tool",
"default": "tool",
"description": "The model will use the specified tool.",
"title": "Type",
"type": "string"
},
"name": {
"description": "The name of the tool to use.",
"title": "Name",
"type": "string"
},
"disable_parallel_tool_use": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"description": "Whether to disable parallel tool use. If true, the model will output exactly one tool use.",
"title": "Disable Parallel Tool Use"
}
},
"required": [
"name"
],
"title": "ToolChoiceToolParam",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-field
disable_parallel_tool_use: bool | None = None
Whether to disable parallel tool use. If true, the model will output exactly one tool use.
pydantic-model
None tool choice — model will not use tools.
Show JSON schema:
Config:
extra:ignore
Fields:
Thinking Configuration
pydantic-model
Extended thinking enabled configuration.
Show JSON schema:
{
"description": "Extended thinking enabled configuration.",
"properties": {
"type": {
"const": "enabled",
"default": "enabled",
"description": "Thinking is enabled.",
"title": "Type",
"type": "string"
},
"budget_tokens": {
"description": "The maximum number of tokens to use for thinking. Minimum is 1024.",
"title": "Budget Tokens",
"type": "integer"
}
},
"required": [
"budget_tokens"
],
"title": "ThinkingConfigEnabledParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
type(Literal['enabled']) -
budget_tokens(int)
pydantic-model
Extended thinking disabled configuration.
Show JSON schema:
Config:
extra:ignore
Fields:
Output Configuration
pydantic-model
Output configuration for structured responses.
Show JSON schema:
{
"$defs": {
"JSONOutputFormatParam": {
"additionalProperties": false,
"description": "JSON output format with schema.",
"properties": {
"type": {
"const": "json_schema",
"default": "json_schema",
"description": "The type of the output format.",
"title": "Type",
"type": "string"
},
"schema": {
"additionalProperties": true,
"description": "The JSON schema of the format.",
"title": "Schema",
"type": "object"
}
},
"required": [
"schema"
],
"title": "JSONOutputFormatParam",
"type": "object"
}
},
"description": "Output configuration for structured responses.",
"properties": {
"format": {
"anyOf": [
{
"$ref": "#/$defs/JSONOutputFormatParam"
},
{
"type": "null"
}
],
"default": null,
"description": "A schema to specify Claude's output format in responses."
}
},
"title": "OutputConfigParam",
"type": "object"
}
Config:
extra:ignore
Fields:
-
format(JSONOutputFormatParam | None)
pydantic-field
format: JSONOutputFormatParam | None = None
A schema to specify Claude's output format in responses.
pydantic-model
JSON output format with schema.
Show JSON schema:
{
"additionalProperties": false,
"description": "JSON output format with schema.",
"properties": {
"type": {
"const": "json_schema",
"default": "json_schema",
"description": "The type of the output format.",
"title": "Type",
"type": "string"
},
"schema": {
"additionalProperties": true,
"description": "The JSON schema of the format.",
"title": "Schema",
"type": "object"
}
},
"required": [
"schema"
],
"title": "JSONOutputFormatParam",
"type": "object"
}
Config:
extra:forbidpopulate_by_name:True
Fields:
Response Types
pydantic-model
Anthropic Messages API response.
This is a Pydantic model representation of the Anthropic Messages API response, returned by Sequrity's Control API when using the Messages format.
Show JSON schema:
{
"$defs": {
"CacheCreation": {
"additionalProperties": true,
"description": "Breakdown of cached tokens by TTL.",
"properties": {},
"title": "CacheCreation",
"type": "object"
},
"RedactedThinkingBlock": {
"additionalProperties": true,
"description": "Redacted thinking content block.",
"properties": {
"type": {
"const": "redacted_thinking",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
}
},
"required": [
"type"
],
"title": "RedactedThinkingBlock",
"type": "object"
},
"ServerToolUsage": {
"additionalProperties": true,
"description": "The number of server tool requests.",
"properties": {},
"title": "ServerToolUsage",
"type": "object"
},
"ServerToolUseBlock": {
"additionalProperties": true,
"description": "Server tool use content block (e.g., bash, web search).",
"properties": {
"type": {
"const": "server_tool_use",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"id": {
"description": "A unique identifier for this server tool use block.",
"title": "Id",
"type": "string"
},
"name": {
"description": "The name of the server tool being used.",
"title": "Name",
"type": "string"
}
},
"required": [
"type",
"id",
"name"
],
"title": "ServerToolUseBlock",
"type": "object"
},
"TextBlock": {
"description": "Text content block in the response.",
"properties": {
"type": {
"const": "text",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"citations": {
"anyOf": [
{
"items": {
"$ref": "#/$defs/TextCitation"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"type",
"text"
],
"title": "TextBlock",
"type": "object"
},
"TextCitation": {
"additionalProperties": true,
"description": "Citation supporting a text block.",
"properties": {
"type": {
"description": "The type of citation (e.g., page_location, char_location).",
"title": "Type",
"type": "string"
}
},
"required": [
"type"
],
"title": "TextCitation",
"type": "object"
},
"ThinkingBlock": {
"description": "Thinking content block showing the model's reasoning process.",
"properties": {
"type": {
"const": "thinking",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"text": {
"description": "The thinking content.",
"title": "Text",
"type": "string"
}
},
"required": [
"type",
"text"
],
"title": "ThinkingBlock",
"type": "object"
},
"ToolUseBlock": {
"description": "Tool use content block \u2014 the model wants to call a tool.",
"properties": {
"type": {
"const": "tool_use",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"id": {
"description": "A unique identifier for this particular tool use block.",
"title": "Id",
"type": "string"
},
"name": {
"description": "The name of the tool being used.",
"title": "Name",
"type": "string"
},
"input": {
"additionalProperties": true,
"description": "An object containing the input being passed to the tool.",
"title": "Input",
"type": "object"
}
},
"required": [
"type",
"id",
"name",
"input"
],
"title": "ToolUseBlock",
"type": "object"
},
"Usage": {
"description": "Token usage statistics for the request.",
"properties": {
"input_tokens": {
"description": "The number of input tokens which were used.",
"title": "Input Tokens",
"type": "integer"
},
"output_tokens": {
"description": "The number of output tokens which were used.",
"title": "Output Tokens",
"type": "integer"
},
"cache_creation_input_tokens": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of input tokens used to create the cache entry.",
"title": "Cache Creation Input Tokens"
},
"cache_read_input_tokens": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of input tokens read from the cache.",
"title": "Cache Read Input Tokens"
},
"cache_creation": {
"anyOf": [
{
"$ref": "#/$defs/CacheCreation"
},
{
"type": "null"
}
],
"default": null,
"description": "Breakdown of cached tokens by TTL."
},
"server_tool_use": {
"anyOf": [
{
"$ref": "#/$defs/ServerToolUsage"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of server tool requests."
},
"service_tier": {
"anyOf": [
{
"enum": [
"standard",
"priority",
"batch"
],
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "If the request used the priority, standard, or batch tier.",
"title": "Service Tier"
}
},
"required": [
"input_tokens",
"output_tokens"
],
"title": "Usage",
"type": "object"
},
"WebSearchToolResultBlock": {
"additionalProperties": true,
"description": "Web search tool result content block.",
"properties": {
"type": {
"const": "web_search_tool_result",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
}
},
"required": [
"type"
],
"title": "WebSearchToolResultBlock",
"type": "object"
}
},
"description": "Anthropic Messages API response.\n\nThis is a Pydantic model representation of the Anthropic Messages API response,\nreturned by Sequrity's Control API when using the Messages format.",
"properties": {
"id": {
"description": "Unique object identifier.",
"title": "Id",
"type": "string"
},
"type": {
"const": "message",
"description": "Object type. Always 'message'.",
"title": "Type",
"type": "string"
},
"role": {
"const": "assistant",
"description": "Conversational role of the generated message. Always 'assistant'.",
"title": "Role",
"type": "string"
},
"content": {
"description": "Content generated by the model as an array of content blocks.",
"items": {
"discriminator": {
"mapping": {
"redacted_thinking": "#/$defs/RedactedThinkingBlock",
"server_tool_use": "#/$defs/ServerToolUseBlock",
"text": "#/$defs/TextBlock",
"thinking": "#/$defs/ThinkingBlock",
"tool_use": "#/$defs/ToolUseBlock",
"web_search_tool_result": "#/$defs/WebSearchToolResultBlock"
},
"propertyName": "type"
},
"oneOf": [
{
"$ref": "#/$defs/TextBlock"
},
{
"$ref": "#/$defs/ThinkingBlock"
},
{
"$ref": "#/$defs/RedactedThinkingBlock"
},
{
"$ref": "#/$defs/ToolUseBlock"
},
{
"$ref": "#/$defs/ServerToolUseBlock"
},
{
"$ref": "#/$defs/WebSearchToolResultBlock"
}
]
},
"title": "Content",
"type": "array"
},
"model": {
"description": "The model that handled the request.",
"title": "Model",
"type": "string"
},
"stop_reason": {
"anyOf": [
{
"enum": [
"end_turn",
"max_tokens",
"stop_sequence",
"tool_use",
"pause_turn",
"refusal"
],
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "The reason the model stopped generating.",
"title": "Stop Reason"
},
"stop_sequence": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Which custom stop sequence was generated, if any.",
"title": "Stop Sequence"
},
"usage": {
"$ref": "#/$defs/Usage",
"description": "Billing and rate-limit usage statistics."
},
"session_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "The Sequrity session ID associated with this response.",
"title": "Session Id"
}
},
"required": [
"id",
"type",
"role",
"content",
"model",
"usage"
],
"title": "AnthropicMessageResponse",
"type": "object"
}
Config:
extra:ignore
Fields:
-
id(str) -
type(Literal['message']) -
role(Literal['assistant']) -
content(list[ContentBlock]) -
model(str) -
stop_reason(Literal['end_turn', 'max_tokens', 'stop_sequence', 'tool_use', 'pause_turn', 'refusal'] | None) -
stop_sequence(str | None) -
usage(Usage) -
session_id(str | None)
pydantic-field
role: Literal['assistant']
Conversational role of the generated message. Always 'assistant'.
pydantic-field
content: list[ContentBlock]
Content generated by the model as an array of content blocks.
pydantic-field
stop_reason: (
Literal[
"end_turn", "max_tokens", "stop_sequence", "tool_use", "pause_turn", "refusal"
]
| None
) = None
The reason the model stopped generating.
pydantic-field
stop_sequence: str | None = None
Which custom stop sequence was generated, if any.
pydantic-field
session_id: str | None = None
The Sequrity session ID associated with this response.
pydantic-model
Text content block in the response.
Show JSON schema:
{
"$defs": {
"TextCitation": {
"additionalProperties": true,
"description": "Citation supporting a text block.",
"properties": {
"type": {
"description": "The type of citation (e.g., page_location, char_location).",
"title": "Type",
"type": "string"
}
},
"required": [
"type"
],
"title": "TextCitation",
"type": "object"
}
},
"description": "Text content block in the response.",
"properties": {
"type": {
"const": "text",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"text": {
"description": "The text content.",
"title": "Text",
"type": "string"
},
"citations": {
"anyOf": [
{
"items": {
"$ref": "#/$defs/TextCitation"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Citations supporting the text block.",
"title": "Citations"
}
},
"required": [
"type",
"text"
],
"title": "TextBlock",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-model
Thinking content block showing the model's reasoning process.
Show JSON schema:
{
"description": "Thinking content block showing the model's reasoning process.",
"properties": {
"type": {
"const": "thinking",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"text": {
"description": "The thinking content.",
"title": "Text",
"type": "string"
}
},
"required": [
"type",
"text"
],
"title": "ThinkingBlock",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-model
Tool use content block — the model wants to call a tool.
Show JSON schema:
{
"description": "Tool use content block \u2014 the model wants to call a tool.",
"properties": {
"type": {
"const": "tool_use",
"description": "The type of the content block.",
"title": "Type",
"type": "string"
},
"id": {
"description": "A unique identifier for this particular tool use block.",
"title": "Id",
"type": "string"
},
"name": {
"description": "The name of the tool being used.",
"title": "Name",
"type": "string"
},
"input": {
"additionalProperties": true,
"description": "An object containing the input being passed to the tool.",
"title": "Input",
"type": "object"
}
},
"required": [
"type",
"id",
"name",
"input"
],
"title": "ToolUseBlock",
"type": "object"
}
Config:
extra:ignore
Fields:
pydantic-model
Token usage statistics for the request.
Show JSON schema:
{
"$defs": {
"CacheCreation": {
"additionalProperties": true,
"description": "Breakdown of cached tokens by TTL.",
"properties": {},
"title": "CacheCreation",
"type": "object"
},
"ServerToolUsage": {
"additionalProperties": true,
"description": "The number of server tool requests.",
"properties": {},
"title": "ServerToolUsage",
"type": "object"
}
},
"description": "Token usage statistics for the request.",
"properties": {
"input_tokens": {
"description": "The number of input tokens which were used.",
"title": "Input Tokens",
"type": "integer"
},
"output_tokens": {
"description": "The number of output tokens which were used.",
"title": "Output Tokens",
"type": "integer"
},
"cache_creation_input_tokens": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of input tokens used to create the cache entry.",
"title": "Cache Creation Input Tokens"
},
"cache_read_input_tokens": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of input tokens read from the cache.",
"title": "Cache Read Input Tokens"
},
"cache_creation": {
"anyOf": [
{
"$ref": "#/$defs/CacheCreation"
},
{
"type": "null"
}
],
"default": null,
"description": "Breakdown of cached tokens by TTL."
},
"server_tool_use": {
"anyOf": [
{
"$ref": "#/$defs/ServerToolUsage"
},
{
"type": "null"
}
],
"default": null,
"description": "The number of server tool requests."
},
"service_tier": {
"anyOf": [
{
"enum": [
"standard",
"priority",
"batch"
],
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "If the request used the priority, standard, or batch tier.",
"title": "Service Tier"
}
},
"required": [
"input_tokens",
"output_tokens"
],
"title": "Usage",
"type": "object"
}
Config:
extra:ignore
Fields:
-
input_tokens(int) -
output_tokens(int) -
cache_creation_input_tokens(int | None) -
cache_read_input_tokens(int | None) -
cache_creation(CacheCreation | None) -
server_tool_use(ServerToolUsage | None) -
service_tier(Literal['standard', 'priority', 'batch'] | None)
pydantic-field
cache_creation_input_tokens: int | None = None
The number of input tokens used to create the cache entry.
pydantic-field
cache_read_input_tokens: int | None = None
The number of input tokens read from the cache.
pydantic-field
Breakdown of cached tokens by TTL.
pydantic-field
The number of server tool requests.
pydantic-field
service_tier: Literal['standard', 'priority', 'batch'] | None = None
If the request used the priority, standard, or batch tier.