laya-mlx
mizorewww/laya-mlx
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
- Stars
- 7,104
- Forks
- 576
Why it is listed
The README presents an independent MLX runtime for open-weight typed decisions on Apple Silicon, relevant as a local Jev-adjacent alternative; it does not claim to be TypeSafe Jev.