Get started

One API, every frontier model, intelligently routed.

Anajak Lab routes each request to the model most likely to nail it — drawing on a pool of the strongest frontier models — behind a single endpoint that speaks both the OpenAI and Anthropic formats. No lock-in, no juggling keys.

What Anajak Lab is

A multi-model gateway, delivered as one API. Instead of committing to a single vendor, Anajak Lab sits in front of a pool of the best models and picks the right one per request. You get one key and one base URL — every SDK and tool you already use keeps working.

Two wire formats. Use the OpenAI format at /v1/chat/completions and /v1/responses, or the Anthropic format at /v1/messages.
Smart routing. Send model: "auto" and Anajak Lab selects the strongest model for the job, or pin a specific model id.
One key, metered. Track spend and remaining credit from /v1/usage — no per-provider accounts.

Quickstart

Grab a key from the dashboard, then change one line — your base URL. Here it is with plain curl:

curl https://api.anajaklabs.dev/v1/chat/completions \
  -H "Authorization: Bearer sk-xxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

Base URL. OpenAI clients use https://api.anajaklabs.dev/v1. Claude Code and the Anthropic SDK use https://api.anajaklabs.dev (no /v1 — the SDK adds it).

Pick your path

Two guided setups, then the full API reference when you need the details: