How to use it
- Search or filter models by licence, size and benchmarks.
- Open a model to compare its details.
- Use the VRAM wizard to check it fits your GPU before you download anything.
How it's built
A local-first, static TypeScript app driven entirely by markdown, so contributors add or update models with a simple pull request.
Why it matters for businesses
Picking the wrong model, or buying the wrong hardware for it, is costly. Comparing licence, quality and memory needs up front avoids both.
In real life
A startup planning to self-host an LLM can confirm in minutes which open models fit its GPUs and allow commercial use.

