OpenAI-compatible - Every model - Auto-routed

One API key.
Every AI model.

Tokaroo routes every request to the best model at the lowest cost - automatically. Drop-in replacement for the OpenAI SDK.

Start freeRead the docs
import { Tokaroo } from "tokaroo";

const client = new Tokaroo({ apiKey: "tok_..." });

// auto - Tokaroo decides everything
// fast - prioritize speed
// max  - maximum capability

const res = await client.chat.completions.create({
  model:    "auto",
  messages: [{ role: "user", content: "Hello!" }],
});

Or swap one line in your existing code:

// Before
const openai = new OpenAI({ apiKey: process.env.OPENAI_KEY });

// After - one line change
const openai = new OpenAI({
  baseURL: "https://api.tokaroo.com/v1",
  apiKey:  process.env.TOKAROO_KEY,
});
Three models. That's it.
auto optimizes for savings. fast prioritizes speed. max buys maximum capability. You pick the tradeoff - we handle everything else.
Open source core
Self-host the routing layer with your own API keys. Free forever. MIT licensed.
Local models
Connect your own Ollama or vLLM server. Routes between cloud and local automatically.
Optimization engine
Cloud adds a scoring engine plus semantic cache - the part that makes it really smart.
Streaming
Full SSE streaming support. Drop-in compatible with any OpenAI client.
Always current
Model pool updated weekly by AI research. Knows which models exist and what they cost.