It speaks the OpenAI, Anthropic and ElevenLabs APIs, so code written against a paid endpoint often works with a one-line change — and it runs without a GPU. That's the whole pitch: keep your app, delete the invoice.
Local models are smaller than the frontier ones. On hard reasoning you'll feel the gap — use it for volume work and keep a cloud model for the hard 5%.
| What you get free | Free and open source (MIT). Drop-in API compatibility with OpenAI, Anthropic and ElevenLabs — no GPU required |
| Where the limit sits | No service limits; bounded by your own hardware |
| Credit card to start | No |
| Pricing type | Open source |
| Last checked | 2026-07-30 · source |
| Run it yourself | Yes — MIT · repo |
| Platforms | windows, mac, linux, api |
| Category | Self-hosted / local · Code & dev · Agents & automation |
We recheck every entry. Subscribers hear first when a free tier quietly shrinks.