CheckerWhat fits my GPU?ModelsWhat to buyEmbeddingsData & methodSubmitAPISponsor

API keys & higher limits: waitlist

Both are live and free: the public read API v1 (no key, 120 requests/minute) and the one-command auto-benchmark. Join the waitlist if you want API keys, higher limits, Hugging Face GGUF lookups via the API, or other paid features later. No dates or prices yet.

1. Will It Run? API FREE v1 LIVE

Fit, offload, KV cache, best quant and measured/estimated speed from your own code. Read the docs. Keys and higher limits: join below.

2. One-command auto-benchmark LIVE

Detects your hardware, runs llama-bench, shows exactly what it would send and asks before submitting. Try it.

We'll only email you about these two things. Your answer about price is only used to decide what to build, not a commitment.