RouterLab

One production entry point

One LLM API for production model routing

Connect applications, agents and workflows to a stable API layer, then choose the model route that fits each workload. RouterLab keeps access, spend visibility and the public model catalog in one place.

Start with RouterLab Browse live model routes

API guides

Start with the protocol. Choose a model family or a client integration only after that.

RouterLab guide map

What you can do

Understand the connection path first, then choose a model from the live catalog.

A stable client contract

Keep one base URL while the selected model changes behind an explicit RouterLab model ID.

Model choice without client rewrites

Use compatible request formats across proprietary and open model families published in the live catalog.

Usage connected to access

Keys, team-level runtime budgets, credits and usage reporting stay attached to the same control layer.

Use an LLM gateway when the interface should outlive the model choice

A direct provider integration is simple for a single fixed model. A gateway becomes useful when an application needs several model families, a common access policy, or the ability to move workloads without changing every client.

OpenAI SDK through the RouterLab LLM API
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ROUTERLAB_API_KEY,
  baseURL: "https://api.routerlab.ch/v1",
});

const response = await client.chat.completions.create({
  model: "model-id-from-/v1/models",
  messages: [{ role: "user", content: "Hello RouterLab" }],
});
Use a model ID returned by GET /v1/models.

Where it fits

Start from the workload and client contract, then choose the route.

See how RouterLab works

Product APIs

Serve chat, extraction, classification and tool-calling features through a common application integration.

Agent fleets

Give coding and autonomous agents a controlled model catalog rather than separate provider credentials.

RAG pipelines

Combine generation routes with embedding models while keeping usage in one operational view.

Get started

  1. 01

    Create an API key

    Open a RouterLab account and create a key for your application or environment.

  2. 02

    Read the live catalog

    Use GET /v1/models and select a published model ID instead of hard-coding a marketing name.

  3. 03

    Change one connection setting

    Point your SDK or client to RouterLab, then keep the rest of your request flow familiar.

Common questions

Use the live catalog and API documentation as the final source for route availability.

What is an LLM API gateway?

It is a stable API layer between an application and one or more model providers. The client calls the gateway, and the gateway applies the published route, access and budget configuration.

Does RouterLab replace the OpenAI SDK?

No. OpenAI-compatible routes are designed to work with the existing SDK by changing the API key, base URL and model ID.

Can the same account use chat and embeddings?

Yes. The public catalog identifies chat and embedding routes, and each request uses the endpoint and model ID appropriate to that workload.

Continue

OpenAI-compatible APIOpen guide Claude Messages APIOpen guide Swiss and European AI APIOpen guide OpenRouter alternativeOpen guide