Chat Completions (OpenAI-compatible)

OpenAI-compatible chat completions endpoint. Accepts the standard OpenAI request schema and returns the standard OpenAI response schema. Supports both streaming (SSE) and non-streaming modes. Works as a drop-in replacement with the OpenAI Python and TypeScript SDKs.

POST /v1/chat/completions

Chat Completions (OpenAI-compatible)

cURL

curl --request POST \
  --url https://api.swarms.world/v1/chat/completions \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: <x-api-key>' \
  --data '
{
  "model": "<string>",
  "messages": [
    {
      "role": "<string>",
      "content": "<string>",
      "name": "<string>"
    }
  ],
  "temperature": 0.5,
  "top_p": 123,
  "n": 1,
  "stream": false,
  "max_tokens": 123,
  "max_completion_tokens": 123,
  "presence_penalty": 123,
  "frequency_penalty": 123,
  "user": "<string>",
  "max_loops": 1
}
'

200

{
  "id": "<string>",
  "created": 123,
  "model": "<string>",
  "choices": [
    {
      "message": {
        "role": "<string>",
        "content": "<string>",
        "name": "<string>"
      },
      "index": 0,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 0,
    "completion_tokens": 0,
    "total_tokens": 0
  },
  "object": "chat.completion"
}

Headers

x-api-key · string · required

Body

application/json

model · string · required

Model to use for completion

messages · ChatMessage · object[] · required

A list of messages comprising the conversation

Show child attributes

temperature · number | null · default: 0.5

Sampling temperature

top_p · number | null

Nucleus sampling parameter

n · integer | null · default: 1

Number of completions to generate

stream · boolean | null · default: false

Whether to stream partial results

max_tokens · integer | null

Maximum tokens to generate

max_completion_tokens · integer | null

Maximum completion tokens (takes precedence over max_tokens)

presence_penalty · number | null

Presence penalty

frequency_penalty · number | null

Frequency penalty

user · string | null

Unique identifier for the end-user

max_loops · integer | null · default: 1

Maximum number of agent reasoning loops (Swarms extension). Defaults to 1 (single pass). Use via extra_body in the OpenAI SDK.

Response

200

application/json

Successful Response

id · string · required

created · integer · required

model · string · required

choices · ChatCompletionChoice · object[] · required

Show child attributes

usage · CompletionUsage · object · required

Show child attributes

object · string · default: chat.completion