Compatible client SDKs

Status: OpenAI SDK is stable for documented endpoints. Other clients are candidates until tested.

NexoRouter does not publish a first-party SDK yet. The recommended integration is direct HTTP or an OpenAI client that lets you configure base_url, API key, and model ID.

Choose a client

“Compatible” does not mean every SDK method is available. NexoRouter only guarantees endpoints published in the API Reference.

Shared configuration

API key: a key created in NexoRouter
Base URL: https://api.nexorouter.com/v1
Model: a current ID from Models or GET /v1/models
Method: Chat Completions

Keep the key on the server. Do not put it in browser-delivered JavaScript, distributed mobile apps, repositories, screenshots, or logs.

Python

Install the client:

python -m pip install --upgrade openai
export NEXOROUTER_API_KEY="your_nexorouter_key"

Create main.py:

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["NEXOROUTER_API_KEY"],
    base_url="https://api.nexorouter.com/v1",
    timeout=60.0,
    max_retries=2,
)

response = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[
        {"role": "system", "content": "Answer clearly and briefly."},
        {"role": "user", "content": "Give me three checks before releasing an API integration."},
    ],
    max_tokens=256,
)

print(response.choices[0].message.content)
print(response.usage)

Run it:

python main.py

Node.js and TypeScript

Install the client:

npm install openai
export NEXOROUTER_API_KEY="your_nexorouter_key"

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.NEXOROUTER_API_KEY,
  baseURL: "https://api.nexorouter.com/v1",
  timeout: 60_000,
  maxRetries: 2,
});

const response = await client.chat.completions.create({
  model: "deepseek-v4-flash",
  messages: [
    { role: "system", content: "Answer clearly and briefly." },
    { role: "user", content: "Give me three checks before releasing an API integration." },
  ],
  max_tokens: 256,
});

console.log(response.choices[0].message.content);
console.log(response.usage);

Direct HTTP with fetch

This option avoids relying on additional SDK methods:

const response = await fetch("https://api.nexorouter.com/v1/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.NEXOROUTER_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "deepseek-v4-flash",
    messages: [{ role: "user", content: "Reply with: connected" }],
    max_tokens: 32,
  }),
});

const body = await response.json();

if (!response.ok) {
  throw new Error(`${response.status} ${body?.error?.code ?? "api_error"}: ${body?.error?.message ?? "Request failed"}`);
}

console.log(body.choices[0].message.content);

Compatibility checks

Before adopting another SDK, confirm that it:

1. Lets you change the base URL.
2. Sends Authorization: Bearer ....
3. Uses POST /v1/chat/completions, not only /v1/responses.
4. Sends the exact model ID without renaming it.
5. Works without streaming.
6. Exposes HTTP status, error body, and rate-limit headers.

If the SDK creates embeddings, Responses API, Anthropic Messages, or Gemini-native requests, that part is outside the current compatibility boundary.

Production

• Create a separate key for each environment and service.
• Set budget, expiry, and model scope.
• Use a timeout and bounded retries only for transient failures.
• Respect retry-after on 429 responses.
• Log request ID, model ID, and status; never log the key or full prompts by default.
• Check Usage Logs for tokens, cost, latency, and errors.

Next

• Quickstart
• API Reference
• Agent integration
• Build a chat app