Why this site exists
Local models keep getting better, but the fun ways to use them are scattered across forums and group chats. I wanted one place to collect them.
Local models have improved fast. A laptop can run a decent chat model, and a single consumer GPU can power a knowledge base, a voice assistant or code completion. But the interesting ways to use them are scattered across Reddit, GitHub and group chats, and they get lost within days.
So this site:
- Plays: uses I’ve collected and (where possible) tested, with the hardware and tools they need
- News: AI filters local-deployment news for me every day, and I review it before it goes up
- Blog: my own experiments
Everything is bilingual, and AI crawlers are welcome.