// docs

Start here.

llmrelay speaks the OpenAI HTTP protocol. If your code already talks to OpenAI, Anthropic, Google or a compatible SDK — you'll be running in two lines.

Endpoint

POST https://api.llmrelay.dev/v1/chat/completions
Authorization: Bearer llm_your_api_key
Content-Type: application/json

All OpenAI-compatible routes are supported: /v1/chat/completions, /v1/completions, /v1/embeddings, /v1/models.

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.llmrelay.dev/v1",
    api_key="llm_...",
)

resp = client.chat.completions.create(
    model="claude-opus-4-7",
    messages=[{"role": "user", "content": "Hello"}],
)

Node.js

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.llmrelay.dev/v1",
  apiKey: "llm_...",
});

const resp = await client.chat.completions.create({
  model: "claude-opus-4-7",
  messages: [{ role: "user", content: "Hello" }],
});

curl

curl https://api.llmrelay.dev/v1/chat/completions \
  -H "Authorization: Bearer llm_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-opus-4-7",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

More endpoints

GET /v1/models — list available models
GET /v1/balance — check remaining credit
POST /v1/embeddings — generate embeddings
POST /v1/images/generations — image generation (select models)

Full reference coming with the public launch. For early access questions: support@llmrelay.dev.