Migrating to AI Token Router

Sign in and create one from the dashboard. It is shown once — we store a hash, so a lost key has to be replaced rather than recovered.

Dashboard → API keys

Two lines. Nothing else in your integration changes.

from openai import OpenAI

client = OpenAI(
    base_url="https://router.xark.io/api/v1",
    api_key="sk-...",            # the key you just created
)

response = client.chat.completions.create(
    model="moonshotai/kimi-k2.6",
    messages=[{"role": "user", "content": "Hello"}],
)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://router.xark.io/api/v1",
  apiKey: process.env.AI_TOKEN_ROUTER_KEY,
});

Anything built on the OpenAI SDK inherits this — LangChain, LlamaIndex, Vercel AI SDK, Instructor. Set the base URL where that framework takes it and the rest is unchanged.

Model names are namespaced publisher/model. The bare name is accepted too, so glm-5.2 and z-ai/glm-5.2 both resolve — use the full id in anything you keep.

Full catalogue — or fetch it as JSON from GET /api/v1/models, which is unauthenticated.

Run one request against the new base URL and compare the response shape to what you had. The usage block is where a billing surprise would first show up, so read it.

curl https://router.xark.io/api/v1/chat/completions \
  -H "Authorization: Bearer sk-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "moonshotai/kimi-k2.6",
    "messages": [{"role": "user", "content": "Say hello"}]
  }'