Text generation¶

POST /api/v5/chat/completions

Generates text. This endpoint is based on the OpenAI-compatible Chat Completions protocol. You submit a model name and the conversation messages and choose between a streaming (SSE) and a non-streaming response; temperature and sampling parameters control generation, and the response returns the generated content together with token usage statistics.

Supported access tokens¶

AtomGit user access tokens

Anonymous access allowed¶

This endpoint does not allow anonymous access.

Request parameters¶

Parameter	Location	Type	Required	Description
`Authorization`	header	string	Yes	User personal token; pass it as `Bearer <token>`

Request Body:

The request body is an OpenAI-compatible Chat Completions request JSON with the following fields:

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "model": {
      "description": "Model name",
      "type": "string"
    },
    "messages": {
      "description": "List of conversation messages",
      "type": "array",
      "items": {
        "type": "string"
      }
    },
    "stream": {
      "description": "Whether to stream the response (SSE). When true, the response is text/event-stream; otherwise it is application/json",
      "type": "boolean",
      "default": false
    },
    "temperature": {
      "description": "Temperature",
      "type": "number",
      "default": 0.7
    },
    "top_p": {
      "description": "top_p sampling parameter",
      "type": "number",
      "default": 0.7
    },
    "top_k": {
      "description": "top_k sampling parameter",
      "type": "integer",
      "default": 50
    },
    "frequency_penalty": {
      "description": "Frequency penalty",
      "type": "number",
      "default": 1.0
    },
    "maxTokens": {
      "description": "Maximum number of tokens",
      "type": "integer"
    }
  },
  "required": ["model", "messages"]
}

Response parameters¶

Field	Type	Always returned	Nullable	Description
`trace_id`	string	Yes	No	Unique identifier of the request
`choices`	array	Yes	No	List of generated results
`ext`	string	Yes	Yes	Extension field, usually null
`usage`	object	Yes	No	Token usage statistics

choices element object

Field	Type	Always returned	Nullable	Description
`choices[].index`	integer	Yes	No	Result index
`choices[].finish_reason`	string	No	Yes	Reason the generation finished; may be null
`choices[].delta`	object	Yes	No	Incremental content

choices[].delta object

Field	Type	Always returned	Nullable	Description
`choices[].delta.content`	string	No	Yes	Fragment of the generated content

usage object

Field	Type	Always returned	Nullable	Description
`usage.prompt_tokens`	integer	Yes	No	Number of tokens consumed by the prompt
`usage.completion_tokens`	integer	Yes	No	Number of tokens consumed by the completion
`usage.total_tokens`	integer	Yes	No	Total number of tokens
`usage.think_tokens`	integer	No	Yes	Number of tokens consumed by the reasoning process

Error response¶

This endpoint has no business-level error response; failures are reported through the HTTP status code (see "HTTP status codes").

HTTP status codes¶

Status code	Description
`200`	Success (OK)
`401`	Authentication required (Requires authentication)
`403`	Access forbidden (Forbidden)
`404`	Request path does not exist
`405`	Request method not supported
`429`	Too many requests; rate limit triggered
`500`	Internal system error

Request example¶

Parameter notes¶

The example below is a non-streaming call. Replace <model> with the model name and <messages> with the list of conversation messages. This endpoint does not allow anonymous access, so you must identify the caller with Authorization: Bearer <token> (replace <token> with your AtomGit user access token).

cURL¶

curl -X POST 'https://api.atomgit.com/api/v5/chat/completions' \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json' \
  -H 'Authorization: Bearer <token>' \
  -d '{
    "model": "<model>",
    "messages": ["<message>"],
    "temperature": 0.7,
    "stream": false
  }'

JavaScript¶

// Native fetch (Node.js 18+ / browsers)
const response = await fetch("https://api.atomgit.com/api/v5/chat/completions", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Accept": "application/json",
    "Authorization": "Bearer <token>",
  },
  body: JSON.stringify({
    model: "<model>",
    messages: ["<message>"],
    temperature: 0.7,
    stream: false,
  }),
});

if (!response.ok) {
  throw new Error(`Request failed: ${response.status}`);
}

const data = await response.json();
console.log(data);

Python¶

import requests

resp = requests.post(
    "https://api.atomgit.com/api/v5/chat/completions",
    headers={
        "Content-Type": "application/json",
        "Accept": "application/json",
        "Authorization": "Bearer <token>",
    },
    json={
        "model": "<model>",
        "messages": ["<message>"],
        "temperature": 0.7,
        "stream": False,
    },
    timeout=10,
)
resp.raise_for_status()

print(resp.json())

Response example¶

Status code: 200

{
  "trace_id": "045560649c7c483e80eb196e2bc95d38",
  "choices": [
    {
      "index": 0,
      "finish_reason": null,
      "delta": {
        "content": "This is a streamed content fragment"
      }
    }
  ],
  "ext": null,
  "usage": {
    "prompt_tokens": 11,
    "completion_tokens": 146,
    "total_tokens": 157,
    "think_tokens": null
  }
}

Error codes¶

Error code	Description
`401`	Unauthenticated or invalid credentials (missing or invalid access credentials)
`400`	Invalid parameters or the request cannot be parsed
`404`	Request path does not exist
`405`	Request method not supported
`429`	Too many requests; rate limit triggered
`500`	Internal system error

Markdown source: View on AtomGit·Content from the open-source project gitcode-dev/atomgit-api-doc