GET

/v1/schemas/nvidia/chat/moonshotai/kimi-k2.6

Accept:

kind inputactivity chathash db99a546…extracted Oct 9, 2026version currentsource upstream spec ↗

field	recorded source
schema	Source linked above. upstream-spec Source trace `37c8f5d5b81032e0f4af572e44b10320dc098030addb3cf2806c57d9d6f3174c` { "fetchedAt": 1791535404, "extractorVersion": "1", "verifiedAt": null }

Request body

11 properties · 2 required · 11 $defs

property	type	description	constraints
messages*	NIMLLMChatCompletionMessage[]	A list of messages comprising the conversation so far.	items ≥ 1
model*	string	The model to use.	default: "moonshotai/kimi-k2.6"
chat_template_kwargs	object	Optional kwargs forwarded to the model chat template (e.g. {"thinking": true} / {"thinking": false}).	nullable
tools	ChatCompletionTool[]	Optional OpenAI-compatible tool definitions for function calling.	nullable
tool_choice	enum \| ChatCompletionNamedToolChoice	Optional OpenAI-compatible tool choice. Use `auto`, `none`, `required`, or a named-tool choice object.	nullable
max_tokens	integer	The maximum number of tokens that can be generated.	1 ≤ n ≤ 65536 · nullable
seed	integer	Changing the seed will produce a different response with similar characteristics. Fixing the seed will reproduce the same results if all other parameters are also kept constant.	-9007199254740991 ≤ n ≤ 9007199254740991 · nullable
stream	boolean	If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format) as they become available, with the stream terminated by a `data: [DONE]`	nullable
stream_options	object	Optional OpenAI-compatible stream options, such as `include_usage`.	nullable
temperature	number	What sampling temperature to use, between 0 and 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.	0 ≤ n ≤ 1 · nullable
top_p	number	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. We generally recommend altering this or `temperature` but not both.	0 < n ≤ 1 · nullable

Types

11 $defs

ChatCompletionFunctionobject

name*	string	Function name.	default: "get_current_weather"
description*	string	Function description.	default: "Get the current weather in a gi…
parameters*	ChatCompletionFunctionParameters	JSON Schema object describing the function arguments.	default: {"type":"object","properties":{}…

ChatCompletionFunctionParametersobject

JSON Schema object describing the function arguments.

default: {"type":"object","properties":{}…

ChatCompletionNamedToolChoiceobject

type*	enum		function · default: "function"
function*	object

ChatCompletionToolobject

type*	enum		function · default: "function"
function*	ChatCompletionFunction

ImageURLobject

url*

string

NIMLLMChatCompletionMessageobject

role*	enum	The role of the message's author.	system \| assistant \| user \| tool
content*	string \| NIMVLMChatCompletionContentPartImage \| NIMVLMChatCompletionContentPartText \| NIMVLMChatCompletionContentPartVideo[]	The contents of the message. <br>To pass media (only with role=`user`): <br> - Use content parts. Text is passed with `type`=`text`. <br> - Images are passed with `type`=`image_url`; set `image_url.url` to an image URL or a base64 data URI like `data:image/{fo…	nullable
name	string	Optional participant or tool name.	nullable
tool_call_id	string	Tool call ID that this tool message is responding to.	nullable
tool_calls	object[]	Tool calls generated by an assistant message.	nullable
reasoning_content	string	Optional preserved reasoning content from a previous assistant turn.	nullable

NIMVLMChatCompletionContentPartImageobject

image_url*	ImageURL
type*	enum		image_url

NIMVLMChatCompletionContentPartTextobject

text*	string
type*	enum		text

NIMVLMChatCompletionContentPartVideoobject

video_url*	VideoURL	Video url
start_offset	number	video offset position	0 ≤ n ≤ 9007199254740991 · nullable
duration	number	video inference duration	0 ≤ n ≤ 9007199254740991 · nullable
type*	enum	The type of the content part.	video_url

Roleenum

system | assistant | user | tool

VideoURLobject

url*

string

Validate a payload

before you spend tokens

shell

$ curl -X POST https://modelschemas.com/v1/validate \
    -d '{"provider":"nvidia","endpointId":"moonshotai/kimi-k2.6","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}