Letta · Schema

AnthropicModelSettings

AIAgentsStateful AgentsMemoryMemGPTContinual LearningMCPMulti-AgentRAGOpen Source

Properties

Name	Type	Description
max_output_tokens	integer	The maximum number of tokens the model can generate.
parallel_tool_calls	boolean	Whether to enable parallel tool calling.
provider_type	string	The type of the provider.
temperature	number	The temperature of the model.
thinking	object	The thinking configuration for the model.
response_format	object	The response format for the model.
verbosity	object	Soft control for how verbose model output should be, used for GPT-5 models.
effort	object	Effort level for supported Anthropic models (controls token spending). 'max' is only available on Opus 4.6. Not setting this gives similar performance to 'high'.
strict	boolean	Enable strict mode for tool calling. When true, tool outputs are guaranteed to match JSON schemas.

View JSON Schema on GitHub

JSON Schema

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "#/components/schemas/AnthropicModelSettings",
  "title": "AnthropicModelSettings",
  "properties": {
    "max_output_tokens": {
      "type": "integer",
      "title": "Max Output Tokens",
      "description": "The maximum number of tokens the model can generate.",
      "default": 4096
    },
    "parallel_tool_calls": {
      "type": "boolean",
      "title": "Parallel Tool Calls",
      "description": "Whether to enable parallel tool calling.",
      "default": true
    },
    "provider_type": {
      "type": "string",
      "const": "anthropic",
      "title": "Provider Type",
      "description": "The type of the provider.",
      "default": "anthropic"
    },
    "temperature": {
      "type": "number",
      "title": "Temperature",
      "description": "The temperature of the model.",
      "default": 1
    },
    "thinking": {
      "$ref": "#/components/schemas/AnthropicThinking",
      "description": "The thinking configuration for the model.",
      "default": {
        "type": "enabled",
        "budget_tokens": 1024
      }
    },
    "response_format": {
      "anyOf": [
        {
          "oneOf": [
            {
              "$ref": "#/components/schemas/TextResponseFormat"
            },
            {
              "$ref": "#/components/schemas/JsonSchemaResponseFormat"
            },
            {
              "$ref": "#/components/schemas/JsonObjectResponseFormat"
            }
          ],
          "discriminator": {
            "propertyName": "type",
            "mapping": {
              "json_object": "#/components/schemas/JsonObjectResponseFormat",
              "json_schema": "#/components/schemas/JsonSchemaResponseFormat",
              "text": "#/components/schemas/TextResponseFormat"
            }
          }
        },
        {
          "type": "null"
        }
      ],
      "title": "Response Format",
      "description": "The response format for the model."
    },
    "verbosity": {
      "anyOf": [
        {
          "type": "string",
          "enum": [
            "low",
            "medium",
            "high"
          ]
        },
        {
          "type": "null"
        }
      ],
      "title": "Verbosity",
      "description": "Soft control for how verbose model output should be, used for GPT-5 models."
    },
    "effort": {
      "anyOf": [
        {
          "type": "string",
          "enum": [
            "low",
            "medium",
            "high",
            "max"
          ]
        },
        {
          "type": "null"
        }
      ],
      "title": "Effort",
      "description": "Effort level for supported Anthropic models (controls token spending). 'max' is only available on Opus 4.6. Not setting this gives similar performance to 'high'."
    },
    "strict": {
      "type": "boolean",
      "title": "Strict",
      "description": "Enable strict mode for tool calling. When true, tool outputs are guaranteed to match JSON schemas.",
      "default": false
    }
  },
  "type": "object"
}