GitHub · Runtime ports · Games and demos

afshinm/laya-mps

Macs where memory matters more than milliseconds.

The repository's GitHub card, read on 2026-09-24

What it does

Runs Laya's typed-decisions checkpoint on an Apple GPU through PyTorch's Metal backend, with a local server and a demo that plays Pong and plots response latency for a minute on request.

The repository's own description: “Run Jev-style typed decisions locally on your Mac with low RAM usage and fast responses”

How it works

PyTorch dispatches the model to Metal Performance Shaders, the server binds to the loopback interface and works offline after the first download, and a lower-memory mode trades latency for roughly a third of the resident footprint.

The repository, by the numbers

Stars

Licence

MIT

Stack

Python · PyTorch MPS · Apple Silicon

Last push

2026-09-21

Read from the GitHub API on 2026-09-24. Stars and the last push move daily — quote them with the date, the way we do.

What we checked

the README, read 2026-09-24
the two memory modes and the disk budget it asks for
the licence file

What we did not check. The 32 ms median and the memory figures. Both are the author's, measured on an M5 Pro; our own run of the same weights is a CPU measurement and is not comparable.

Open the repository ↗All builds

The rest of the topic

Jev, the reference →

Start here

Laya, the reference What it is, what we measured running it, and what it is not.
Laya and Jev, compared The two models on the same questions, including the ones Laya loses.

Columns

Built with Laya 18 you are here Ports, servers, agents and games, with what we checked.
What X is saying 22
What people filmed 9
What Hacker News made of it 6
Docs, weights and reading 15

The topic

Every record, by kind of work The same rows cut a second way, across every column.
What self-hosting costs Hardware, memory, and the failure modes of running it yourself.
What the licence allows Apache-2.0 weights, three checkpoints, and the limits that come with them.

By kind of work

Runtime ports 23 Getting the weights onto someone else's hardware: MLX on a Mac, Core ML on the Neural Engine, ggml in C++, candle in Rust, ONNX in Node.
Serving and API 8 Standing the model up as something else can call, usually behind Jev's own wire format.
Agents and tools 6 Typed decisions inside an agent loop: routing, evals, guardrails, memory, and the plumbing around them.
Browser and computer use 2 Deciding the next click, key or field on a real screen.
Games and demos 9 The model playing something you can watch. Most of this model's traffic arrived through one of these.
Benchmarks and evaluation 19 Numbers somebody actually ran — including the ones that say Laya lost.

What it does

How it works

The repository, by the numbers

What we checked

More like this

mizorewww/laya-mlx

mizorewww/laya-coreml

receptron/laya