GET

/v1/schemas/grok/chat/v1/messages

Accept:

kind inputactivity chathash 8e6f9a77…extracted Oct 21, 2026version currentsource upstream spec ↗

field	recorded source
schema	Source linked above. upstream-spec Source trace `6910c22b5fda89bcef4649f7f3798b75198027b07a3553f7ad05ec62a3b2b210` { "fetchedAt": 1792559435, "extractorVersion": "1", "verifiedAt": null }

Request body

12 properties · 0 required · 12 $defs

property	type	description	constraints
max_tokens	integer	The maximum number of tokens to generate before stopping. The model may stop before the max_tokens when it reaches the stop sequence.	format: int32
messages	MessageBody[]	Input messages.
metadata	MessageMetadata	An object describing metadata about the request.	nullable
model	string	Model name for the model to use.
stop_sequences	any	(Not supported by reasoning models) Up to 4 sequences where the API will stop generating further tokens.
stream	any	If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a `data: [DONE]` message.
system	string \| SystemMessagePart[]	System prompt message for the model, defining how the model should behave to user messages.	nullable
temperature	any	What sampling temperature to use, between 0 and 2. Values above 1 are sampled at 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. It may not work well with reasoning models.	0 ≤ n ≤ 2 · format: float · default: 1
tool_choice	object \| object \| object	Controls which (if any) tool is called by the model. `"none"` means the model will not call any tool and instead generates a message. `"auto"` means the model can pick between generating a message or calling one or more tools. `"any"` means the model must call…	nullable
tools	any	A list of tools the model may call in JSON-schema. Currently, only functions are supported as a tool. Use this to provide a list of functions the model may generate JSON inputs for. A max of 1000 functions are supported.
top_k	any	(Unsupported) When generating next tokens, randomly selecting the next token from the k most likely options.	format: int32
top_p	any	An alternative to sampling with `temperature`, called nucleus sampling, where the model considers the results of the tokens with `top_p` probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. It is generally reco…	0 < n ≤ 1 · format: float · default: 0.95

Types

12 $defs

MessageBodyobject

Anthropic compatible message body

content*	string \| MessageContentPart[]	The content message.
role*	string	The role that the message belongs to, `"system"` for system prompt, `"user"` for user prompt, and `"assistant"` for response from the model.

MessageContentstring | MessageContentPart[]

MessageContentPartobject | object | object | …

MessageImageContentobject | object

MessageMetadataobject

user_id

any

A unique identifier representing your end-user, which can help xAI to monitor and detect abuse.

MessageToolChoiceobject | object | object

Tool choice option.

MessageToolInputSchemaobject

properties*	any	JSON-object of the tool input schema.
required	any	Required properties of the tool input schema, if any.
type*	string	Type of the schema. This is always `"object"`.

MessageToolsobject

cache_control	any	(Unsupported) Cache control.
description*	string	Description of the tool.
input_schema*	MessageToolInputSchema	Input schema allowed by the tool.
name*	string	Name of the tool.

SystemMessageContentstring | SystemMessagePart[]

SystemMessagePartobject

cache_control	any	(Unsupported) Cache control.
text*	string	System prompt text.
type*	string	Type of the object. This is always `"text"`.

ToolResultContentstring | ToolResultContentBlock[]

Content of a tool_result block. The Anthropic SDK may send this as either a plain string or an array of typed content blocks (e.g. `[{"type": "text", "text": "..."}]`).

ToolResultContentBlockobject | object

A single content block within a tool_result's array content.

Validate a payload

before you spend tokens

shell

$ curl -X POST https://modelschemas.com/v1/validate \
    -d '{"provider":"grok","endpointId":"v1/messages","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}