llm.buzz, the open market for inference, is public

llm.buzz is our AI Gateway: one API key for every model, with no markup on what providers charge.

A market, not a list

A model name is a slot that several providers can fill. Each provider posts its own price, and the router compares them on every request. You pick how it chooses, per model:

• smart: price and time to first token together. Every new model starts here.
• cheapest or fastest, for one model or, with :floor and :nitro, for one request.
• order: a fixed order of providers.

If the first provider errors or times out, the next one answers, and you pay the price of the provider that served you.

Where your prompts go

Every endpoint declares whether it serves from the EU, whether it trains on prompts and how many days it keeps them. Routing can stay inside the EU on every plan.

Sell your inference

Anyone who serves or fine-tunes models can list an endpoint at their own price. A listing goes live when the endpoint passes an automated check, and leaves routing when it fails one.

How it got here

Work started on 19 August 2026. EU-only routing came on 12 October, prepaid credit on 14 October, and providers began applying and setting their prices on 18 October. On 21 October the rankings and the apps pages followed, with the post A market, not a gateway.