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.
laya-mlx launched on 19 September 2026. It ranked #58 on Daily Launch on 14 October 2026 with a Launch Score of 42, from 6.9k stars in 24 days on GitHub. It was new in the ranking that day.
What is Laya Mlx?
Open-weight typed decisions, running natively on Apple Silicon. 13.4 ms median end-to-end for a short English typed decision. 7.4 ms with the multilingual checkpoint. 0 output tokens. Local MLX inference, with no PyTorch, Transformers runtime, or cloud API. The GIF is an original-speed render of a real local Snake run. Every move calls Laya; the visible cycle safety layer can correct unsafe proposals. The latency figures above are the separate one-question API benchmark, not the frame time of the three-question Snake loop.
From its README on GitHub
Laya Mlx (laya-mlx) is a repo by mizorewww on GitHub. It is written in Python and released under the Apache-2.0 license.
How to get Laya Mlx
pip install laya-mlxFrom its README on GitHub
More steps and the source are in its repository on GitHub.
Launch timeline
Launch Score, last 7 days
Ranked on one day so far: 42 of 100 on 14 October.