START / API

Models and routing

Connect your existing SDKs through Mengbi Router’s unified model entry point.

AUTH / PERSONAL KEY

Request headers

Each protocol uses a different authentication header. Configure the one that matches your endpoint.

ProtocolHeaderNotes
Anthropicx-api-key: sk-mengbi-xxxxxxAlso send anthropic-version: 2023-06-01
OpenAIAuthorization: Bearer sk-mengbi-xxxxxxFor Chat Completions and Responses API
GeminiAuthorization: Bearer sk-mengbi-xxxxxxThe native endpoint also accepts x-goog-api-key
!

This page is the preview integration contract for individual Mengbi Router users. Examples will remain compatible when the production gateway opens.

02 /

Quickstart

Create an API key in your personal console, then keep it in a server-side environment variable. Never put the key in browser code.

Set environment variables

环境变量bash
export MENGBI_ROUTER_API_KEY="sk-mengbi-xxxxxx"
export MENGBI_ROUTER_BASE_URL="https://router-api.mengbi-ai.net/v1"

Notesk-mengbi-xxxxxx is only a placeholder. Replace it with your own personal key.

Your first OpenAI-compatible request

POST /v1/chat/completions
cURLbash
curl https://router-api.mengbi-ai.net/v1/chat/completions \
  -H "Authorization: Bearer sk-mengbi-xxxxxx" \
  -H "content-type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "messages": [{"role": "user", "content": "你好"}]
  }'
Python · openaipython
from openai import OpenAI

client = OpenAI(
    api_key="sk-mengbi-xxxxxx",
    base_url="https://router-api.mengbi-ai.net/v1",
)

response = client.chat.completions.create(
    model="gpt-5.6-sol",
    messages=[{"role": "user", "content": "你好"}],
)
print(response.choices[0].message.content)
03 /

Model list

Call models by ID. The model catalog shows currently available models with published prices; read the endpoint list before integrating.

Read available models

GET /v1/models
cURLbash
curl https://router-api.mengbi-ai.net/v1/models \
  -H "Authorization: Bearer sk-mengbi-xxxxxx"
返回结构json
{
  "object": "list",
  "data": [
    {"id": "gpt-5.6-sol", "object": "model", "owned_by": "mengbi"},
    {"id": "claude-sonnet-5", "object": "model", "owned_by": "anthropic"}
  ]
}

NoteAvailability, model IDs, and prices can change. Do not permanently cache the full model list on the client.

04 /

Anthropic native

Use the native Anthropic protocol for Claude models to preserve message structure, streaming events, and model capabilities.

Basic request

POST /v1/messages
cURLbash
curl https://router-api.mengbi-ai.net/v1/messages \
  -H "x-api-key: sk-mengbi-xxxxxx" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-sonnet-5",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "用一句话介绍你自己"}]
  }'
Python · anthropicpython
from anthropic import Anthropic

client = Anthropic(
    api_key="sk-mengbi-xxxxxx",
    base_url="https://router-api.mengbi-ai.net",
)

response = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "你好"}],
)
print(response.content[0].text)

Streaming request

Add stream: true to the request body. The response uses text/event-stream; keep reading events instead of parsing it as one JSON document.

请求体json
{
  "model": "claude-sonnet-5",
  "max_tokens": 1024,
  "stream": true,
  "messages": [{"role": "user", "content": "写一首五言绝句"}]
}

NoteA common event sequence is message_start, content_block_delta, message_delta, and message_stop.

05 /

OpenAI-compatible

For the OpenAI SDK, Codex, and most clients that support Chat Completions.

Chat Completions

POST /v1/chat/completions
cURLbash
curl https://router-api.mengbi-ai.net/v1/chat/completions \
  -H "Authorization: Bearer sk-mengbi-xxxxxx" \
  -H "content-type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello"}]
  }'
JavaScript · openaijavascript
import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: process.env.MENGBI_ROUTER_API_KEY,
  baseURL: 'https://router-api.mengbi-ai.net/v1',
})

const response = await client.chat.completions.create({
  model: 'gpt-5.6-sol',
  messages: [{ role: 'user', content: '你好' }],
})
console.log(response.choices[0].message.content)

Streaming request

Add stream: true to the Chat Completions request. The response uses standard OpenAI SSE chunks and ends with data: [DONE].

请求体json
{
  "model": "gpt-5.6-sol",
  "stream": true,
  "messages": [{"role": "user", "content": "写一段介绍"}]
}
06 /

Responses API

For applications already using the OpenAI Responses SDK. Currently intended for GPT models.

Basic request

POST /v1/responses
cURLbash
curl https://router-api.mengbi-ai.net/v1/responses \
  -H "Authorization: Bearer sk-mengbi-xxxxxx" \
  -H "content-type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "input": "用一句话介绍你自己"
  }'
Python · openaipython
from openai import OpenAI

client = OpenAI(
    api_key="sk-mengbi-xxxxxx",
    base_url="https://router-api.mengbi-ai.net/v1",
)

response = client.responses.create(
    model="gpt-5.6-sol",
    input="用一句话介绍你自己",
)
print(response.output_text)

NoteUse /v1/messages for Claude and the Gemini native format for Gemini. Sending an incompatible model to /v1/responses returns an error.

Fast mode

If the model catalog marks a model as supporting fast mode, add service_tier: fast to Chat Completions or Responses API requests. Confirm availability from the model catalog and response.

请求字段json
{
  "model": "gpt-5.6-sol",
  "input": "分析这个需求",
  "service_tier": "fast"
}
07 /

Gemini native

Gemini CLI and native Gemini SDKs can use the Gemini-style path and request body.

Generate content

POST /v1beta/models/{model}:generateContent
cURLbash
curl https://router-api.mengbi-ai.net/v1beta/models/gemini-3.6-flash:generateContent \
  -H "Authorization: Bearer sk-mengbi-xxxxxx" \
  -H "content-type: application/json" \
  -d '{
    "contents": [{
      "role": "user",
      "parts": [{"text": "用一句话介绍你自己"}]
    }]
  }'
返回结构json
{
  "candidates": [{
    "content": {"role": "model", "parts": [{"text": "你好！"}]},
    "finishReason": "STOP"
  }],
  "usageMetadata": {
    "promptTokenCount": 12,
    "candidatesTokenCount": 22
  }
}

Streaming and environment variables

Use streamGenerateContent with alt=sse for streaming. Different Gemini clients may use different field names for a custom Base URL.

SSEbash
curl "https://router-api.mengbi-ai.net/v1beta/models/gemini-3.6-flash:streamGenerateContent?alt=sse" \
  -H "Authorization: Bearer sk-mengbi-xxxxxx" \
  -H "content-type: application/json" \
  -d '{"contents": [{"role": "user", "parts": [{"text": "你好"}]}]}'
环境变量bash
export GOOGLE_GEMINI_BASE_URL="https://router-api.mengbi-ai.net"
export GEMINI_API_KEY="sk-mengbi-xxxxxx"
export GEMINI_API_KEY_AUTH_MECHANISM="bearer"