Laya Ultrafast: a local port of the jev-ultrafast browser agent
The browser-use agent's decisions move from the hosted Jev API to laya-mlx, keeping the DOM snapshot, executor and safety checks.
BUILDS
113 builds made with System One models.
Agents, routers, scorers and games — built on models that decide. Every entry links to the models it uses, so you can pull the same one and start from there.
63 builds · page 2 of 3
Laya vs Jev Arena: a Snake race and a fighting game
Both models get identical inputs and questions in unscripted matches, as shown in a Prompt Engineer video; only the model changes.
laya-zig: a dependency-free Zig runtime for CPUs
Tokenizer, encoder and decision head for the 322M multilingual checkpoint in one binary, with no Python, torch or BLAS.
Laya4j: a pure-Java SDK on ONNX Runtime
A Java 17+ port with typed questions, a script-detecting router that picks the right checkpoint, and model downloads from the Hugging Face Hub.
Lunar Laya: a fine-tuned lander that lands 30 of 30
Arcade physics, a browser flight recorder and a Laya checkpoint fine-tuned on CUDA that landed every held-out flight when run on MLX.
pydecide: one Python client with fallback across decision backends
Typed questions against Jev, OpenRouter, Laya or cross-encoders, with a confidence policy that moves to the next backend on errors or low confidence.
Slop Finder: flags formulaic AI-style posts as you scroll
A Chromium extension for X, LinkedIn and Reddit that scores posts locally through a Laya-MLX helper.
Can decision models catch SQL bugs?
Fourteen PostgreSQL statements, half with deliberate bugs such as fan-out sums, = NULL and UPDATE without WHERE, each reviewed with four typed questions.
arbiter: a Jev-compatible decision server for Laya or your own model
A self-hosted typed-decision server for NVIDIA GPUs and Apple Silicon that speaks Jev's wire format, with integrations for coding agents.
docker-laya: a complete Laya API service in one container
A FastAPI service that bundles all three Laya checkpoints behind a language router, with API keys, presets, OpenAPI docs and bulk inference.
EdgeJev: offline Laya on a 4-core CPU in 15.6 ms
A convert, quantize and deploy toolchain that turns Laya and related models into INT8 ONNX for offline CPUs, measured at 15.6 ms per question.
A fully local Home Assistant conversation agent on Laya
Allen Porter's Assist agent resolves intents and entities in a single Laya pass on CPU, CUDA or Apple Silicon, with nothing sent to the cloud.
Reranking Finnish legal passages with typed-judgment models
Compares Jev, open-weight Laya, cross-encoders and GPT models as rerankers over a Postgres full-text and pgvector first stage, with cost per 1,000 queries.
jevals: agent evals and guardrails as typed decisions
Openlayer's library bundles the evaluations for an agent trace into one decision call, cheap enough to run on every trace — or free with local Laya.
laya-candle: Rust inference with custom Metal kernels
Runs the original Laya safetensors through Candle with fused kernels, tiled attention and GEMM dispatch tuned for the M1 Pro.
laya.cpp: a standalone C++ runtime with CUDA and Vulkan
A ggml-based runtime for all three checkpoints with native tokenization, decision heads and JSON output.
laya_ex: Laya in Elixir on Nx and Bumblebee
Fetches the official checkpoint on first use and answers typed questions natively, with EMLX on Apple Silicon.
laya-jev-lab: Laya first, Jev only when unsure
A local-first cascade: confident Laya answers run straight away and uncertain ones escalate to Jev, matching Jev's accuracy at 1.8× the speed.
laya-mcp: laya-mlx for any MCP client
Will Sargent's stdio MCP server lets Claude Code and other clients classify text, score against a rubric or ask yes/no questions locally.
Laya clears Super Mario Bros. 3, World 1-1
Unmodified weights on a DGX Spark, real NES emulation and text observations read from RAM: 207 verified decisions at a 16.79 ms median.
laya-pong: a browser paddle at 18.7 ms per frame
One typed question per frame on Metal, with the opposing paddle driven by three lines of arithmetic for comparison.
laya_router: as accurate as GPT-5 nano at routing, 35× faster
Two routers behind one endpoint on 180 labeled requests; Laya picks small, medium or powerful in a single forward pass, for free.
laya-rust: pure-Rust Laya inference
Loads the safetensors directly with candle and runs any checkpoint on CPU, Metal or CUDA — no Python, torch or ONNX export.
Fine-tuning Laya for real-time Tetris placements
The 421M model learns to choose among enumerated placements scored by a board-evaluation teacher, with a guide to reproduce it.
Many of the first builds here were collected by madewithlaya.com. Every entry links to its creator’s own post, repository or site.