Snowflake · Schema

CompleteRequest

LLM text completion request.

Data LakesData SharingData WarehousingDatabaseSQL

Properties

Name	Type	Description
model	string	The model name. See documentation for possible values.
messages	array
temperature	number	Temperature controls the amount of randomness used in response generation. A higher temperature corresponds to more randomness.
top_p	number	Threshold probability for nucleus sampling. A higher top-p value increases the diversity of tokens that the model considers, while a lower value results in more predictable output.
max_tokens	integer	The maximum number of output tokens to produce. The default value is model-dependent.
max_output_tokens	integer	Deprecated in favor of "max_tokens", which has identical behavior.
response_format	object	An object describing response format config for structured-output mode.
guardrails	object
tools	array	List of tools to be used during tool calling
tool_choice	object
provisioned_throughput_id	string	The provisioned throughput ID to be used with the request.
sf-ml-xp-inflight-prompt-action	string	Reserved
sf-ml-xp-inflight-prompt-client-id	string	Reserved
sf-ml-xp-inflight-prompt-public-key	string	Reserved
stream	boolean	Reserved

View JSON Schema on GitHub

JSON Schema

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "#/components/schemas/CompleteRequest",
  "title": "CompleteRequest",
  "type": "object",
  "description": "LLM text completion request.",
  "properties": {
    "model": {
      "description": "The model name. See documentation for possible values.",
      "type": "string",
      "example": "example_value"
    },
    "messages": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "role": {
            "type": "string",
            "description": "Indicates the role of the message, one of 'system', 'user' or 'assistant'.\n\nRules:\n  - A 'user' message must be the last message in the list.\n  - If a 'system' message is specified, it must be the first message.\n  - If a 'assistant' message is specified, it must be immediately before a 'user' message in the list.\n\nMultiple 'assistant' and 'user' messages can be specified, but they must alternate in sequence.\n",
            "default": "user"
          },
          "content": {
            "type": "string",
            "description": "The text completion prompt, e.g. 'What is a Large Language Model?'."
          },
          "content_list": {
            "type": "array",
            "description": "Contents of toolUse and toolResults",
            "items": {
              "discriminator": {
                "propertyName": "type",
                "mapping": {
                  "text": "common-cortex-tool.yaml#/components/schemas/TextContent",
                  "tool_result": "common-cortex-tool.yaml#/components/schemas/ToolResults",
                  "tool_use": "common-cortex-tool.yaml#/components/schemas/ToolUse"
                }
              }
            }
          }
        },
        "required": [
          "content"
        ]
      },
      "minItems": 1,
      "example": []
    },
    "temperature": {
      "description": "Temperature controls the amount of randomness used in response generation. A higher temperature corresponds to more randomness.",
      "type": "number",
      "nullable": true,
      "minimum": 0.0,
      "example": 42.5
    },
    "top_p": {
      "description": "Threshold probability for nucleus sampling. A higher top-p value increases the diversity of tokens that the model considers, while a lower value results in more predictable output.",
      "type": "number",
      "default": 1.0,
      "minimum": 0.0,
      "maximum": 1.0,
      "example": 42.5
    },
    "max_tokens": {
      "description": "The maximum number of output tokens to produce. The default value is model-dependent.",
      "type": "integer",
      "default": 4096,
      "minimum": 0,
      "example": 10
    },
    "max_output_tokens": {
      "deprecated": true,
      "description": "Deprecated in favor of \"max_tokens\", which has identical behavior.",
      "type": "integer",
      "nullable": true,
      "example": 10
    },
    "response_format": {
      "type": "object",
      "nullable": true,
      "description": "An object describing response format config for structured-output mode.",
      "properties": {
        "type": {
          "type": "string",
          "enum": [
            "json"
          ],
          "description": "The response format type (e.g., \"json\")."
        },
        "schema": {
          "type": "object",
          "description": "The schema defining the structure of the response. If the `type` is \"json\", the `schema` field should contain a valid JSON schema."
        }
      },
      "example": "example_value"
    },
    "guardrails": {
      "$ref": "#/components/schemas/GuardrailsConfig"
    },
    "tools": {
      "description": "List of tools to be used during tool calling",
      "type": "array",
      "items": {
        "$ref": "common-cortex-tool.yaml#/components/schemas/Tool"
      },
      "example": []
    },
    "tool_choice": {
      "$ref": "common-cortex-tool.yaml#/components/schemas/ToolChoice"
    },
    "provisioned_throughput_id": {
      "type": "string",
      "description": "The provisioned throughput ID to be used with the request.",
      "nullable": true,
      "example": "500123"
    },
    "sf-ml-xp-inflight-prompt-action": {
      "type": "string",
      "description": "Reserved",
      "example": "example_value"
    },
    "sf-ml-xp-inflight-prompt-client-id": {
      "type": "string",
      "description": "Reserved",
      "example": "500123"
    },
    "sf-ml-xp-inflight-prompt-public-key": {
      "type": "string",
      "description": "Reserved",
      "example": "example_value"
    },
    "stream": {
      "type": "boolean",
      "default": true,
      "nullable": true,
      "description": "Reserved",
      "example": true
    }
  },
  "required": [
    "model",
    "messages"
  ]
}