Jev and Laya decision models
Jev and Laya are both System One decision models: send state and typed questions, get back structured answers your code can use directly. This site carries the full documentation for Jev, Laya and other decision models, plus a cross-model comparison.
Decision models
Jev
Closed hosted API — the reference implementation of typed decisions
ProprietarychoicescorenoulLaya
Open weights, self-hostable, same API shape as Jev
Apache-2.0choicescorenoulKev
Small Jev-style decision models you can train and run yourself
Apache-2.0choicescorenoulLaya-CoreML
Laya weights in Core ML, on the Neural Engine, with zero generated tokens
Apache-2.0choicescorenoulLaya-MLX
Laya weights ported to MLX, running natively on Apple silicon
Apache-2.0choicescorenoul
Keep exploring
Compare Jev vs Laya vs Kev vs Laya-CoreML vs Laya-MLX
Protocols, primitives and pricing, side by side
Decision model glossary
What choice, score and noul actually mean
Decision model field guide
How these models get used in practice, and where they break
Run models locally
Tools for running open decision models on your own machine
About
Where the content comes from and how it stays current