kimi-k3

文本KimiKimikimi-k3

Open Playground

Kimi

Authentication

Base URL: https://host.docker.internal:8000
Authorization: Authorization: Bearer YOUR_API_KEY
Content-Type: application/json
Model name: kimi-k3

Integration protocols

OpenAI 兼容兼容

Why this protocol

适合已有 OpenAI SDK、Chat Completions 代码或统一网关接入的用户。

Endpoint

POST/v1/chat/completions

Authorization: Bearer YOUR_API_KEYContent-Type: application/json

Request / Response JSON

// REQUEST

{
  "model": "kimi-k3",
  "messages": [
    {
      "content": "你好，世界",
      "role": "user"
    }
  ],
  "temperature": 1,
  "top_p": 0.95,
  "max_tokens": 4096,
  "stream": false,
  "presence_penalty": 0,
  "frequency_penalty": 0,
  "n": 1,
  "thinking": {
    "type": "enabled"
  }
}

// RESPONSE

{
  "id": "chatcmpl_example",
  "model": "gpt-5.5",
  "choices": [
    {
      "message": {
        "content": "你好，世界"
      }
    }
  ],
  "usage": {
    "prompt_tokens": 8,
    "completion_tokens": 16,
    "total_tokens": 24
  }
}

Request / Response schema

// INPUT SCHEMA

Field	Type	Required
messages	array	Y
temperature	number	N
top_p	number	N
max_tokens	integer	N
stream	boolean	N
presence_penalty	number	N
frequency_penalty	number	N
n	number	N
thinking	object	N

// RESPONSE SCHEMA

Field	Type	Required
id	string	N
model	string	N
choices[].message.content	string	N
usage.prompt_tokens	integer	N
usage.completion_tokens	integer	N
usage.total_tokens	integer	N

Code examples

cURLrequest.sh

curl -X POST "https://host.docker.internal:8000/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
  "model": "kimi-k3",
  "messages": [
    {
      "content": "你好，世界",
      "role": "user"
    }
  ],
  "temperature": 1,
  "top_p": 0.95,
  "max_tokens": 4096,
  "stream": false,
  "presence_penalty": 0,
  "frequency_penalty": 0,
  "n": 1,
  "thinking": {
    "type": "enabled"
  }
}'