Last released Nov 19, 2026
Local-first LLM serving for software that has to work: a resident engine, typed errors, validated extraction and an OpenAI-compatible API — on hardware you don't control, with data that can't leave.
Last released Feb 17, 2025
None