Laya alternatives
Jev was released in September 2026, and within weeks several open models with the same typed-question interface appeared. Laya is a 322M to 421M encoder. Most alternatives are open LLMs, mostly Qwen, fine-tuned at 0.5B to 35B parameters to score options without generating text. They trade speed for accuracy.
468 projects
laya-mlx
mizorewww
Runs Laya typed-decision models locally on Apple Silicon using MLX
laya-coreml
mizorewww
Runs Laya typed-decision models locally on Apple Silicon using Core ML and the Neural Engine
openJev-verdict-2.0
Heman10x-NGU
Develops and benchmarks a non-autoregressive typed-decision model against Jev and Laya
laya-playground
wdobry
A local website combines Laya model demos, games, a benchmark, and an agent skill
laya-vs-jev
virajbhartiya
A side-by-side T-Rex game compares local Laya and hosted Jev decision models
laya-mps
afshinm
Runs Laya typed-decision inference locally on Apple Silicon using Metal Performance Shaders
eikos
caiovicentino
Builds calibrated typed-decision models for finance and trading and evaluates them against Laya
JevAny
SimpleJev
Training and evaluation infrastructure for System 1 decision models, with Laya as a benchmark baseline
lev
jlt-commons
Implements a typed decision engine using Laya checkpoints and GGUF chat models
edgejev
yzfly
Exports, quantizes, and serves Laya and other Jev-style models for offline CPU inference
laya-jev-lab
yibie
Compares Jev and Laya decision models and evaluates a local-first inference cascade
opendecider
manjunathshiva
Provides calibrated System 1 decision models and benchmarks them against Laya and other models
laya_router
glukicov
A Python model router compares local Laya decisions with a GPT-5 nano routing model
jevbench
dhruvmehra
A reproducible benchmark compares JEV, Laya, and other classifiers across datasets and metrics
laya-as-judge
rbrus
Uses Laya-MLX to evaluate models and agents with typed-decision judgments
laya-plays-smb3
cv
Demonstrates Laya controlling Super Mario Bros. 3 with recorded, replay-verified decisions
reflexbench
brida-ai
Benchmarks Laya and other typed-decision engines across quality, calibration, robustness, and latency
laya-jev-benchmark
Luni
Benchmarks Laya against Jev and other models on phishing detection and calibration
laya-onnx
MstyAI
Runs Laya decision models locally with ONNX Runtime through a Go library and CLI
jev-tests
schacon
Compares Laya, Jev, Kev, and Claude in three macOS typed-decision demos
MacJev-322M-4K-Laya
chaoliangUNSW
Fine-tunes Laya for long-context local Mac agents and compares it with the base model
jev-laya-benchmark
harrymunro
Benchmarks local MLX Laya against TypeSafe's hosted Jev on synthetic decision tasks
sysone-bench
instax-dutta
Compares Laya and Jev on identical inputs across multiple decision benchmarks
runtime-tutorials
Runtime-weekly
Provides runnable Python guides for Laya classification and a comparison of Laya with other systems
The models
Sizes and licences are from each model's Hugging Face card.
| model | maker | sizes | licence |
|---|---|---|---|
| Kev | Jared Palmer | 0.5B to 27B (benchmarked: 0.8B, 4B, 9B) | Apache 2.0 |
| Decider | Mapika | 0.8B to 35B, plus a 2B vision model | Apache 2.0 |
| Clef, Clef-flash | Cloudflare | 27B, 9B | Apache 2.0 |
| GLiNER2.5-Decide | Fastino | 340M English, 287M multilingual, 1B | Apache 2.0 |
| Julia-1 | Supersonic Labs | 144M | Apache 2.0 |
| Jebadiah | Frontier Infra | 4B, 9B, 27B | Apache 2.0 |
| pplx-decider | Perplexity | ~26B ("27b" in the name) | Apache 2.0 |
- Kev runs a frozen Qwen3.5 base prefill-only, with a LoRA adapter and a pointer head over the options. The card says states of up to 65,536 tokens are served and 8,192 are validated.
- Mapika Decider calls itself an open reproduction of the System One model class. Its training code and data registry are on GitHub.
- Clef and Clef-flash read text, JSON, images and video, per their cards. Cloudflare's announcement says they are Jev-API compatible and also served on Workers AI.
- GLiNER2.5-Decide is a classification-only member of the GLiNER2.5 family: any label set at call time, single- or multi-label.
- Julia-1 is, like Laya, an encoder: it starts from mmBERT-small. See ModernBERT and mmBERT in Laya.
- Jebadiah ships a standalone server that answers
POST /v1/systemoneon one CUDA GPU or an Apple Silicon Mac. - pplx-decider v1.1 uses a separate decision readout instead of a normal language-model head, so its card asks you to use the included inference code.
The hosted reference is Jev from TypeSafe. Its weights are closed and it is billed per input token. The differences from Laya are in Laya vs Jev.
What independent benchmarks found
akitaonrails (October 10, 2026) used synthetic, AI-assisted cases in English and Brazilian Portuguese without independent human labels. Chance was about 31%. Local models ran on a Ryzen AI Max+ 395. Compact-set accuracy and median latency:
| model | accuracy | latency |
|---|---|---|
| Jev (hosted) | 97.6% | 255 ms |
| Clef (hosted) | 95.8% | 302 ms |
| Perplexity Decider 1.1 | 99.4% | 1,547 ms |
| Jebadiah 4B v2 | 90.5% | 205 ms |
| Kev 9B | 88.1% | 561 ms |
| Kev 4B | 87.5% | 246 ms |
| Mapika Decider | 87.5% | 186 ms |
| Clef-flash (local) | 87.5% | 618 ms |
| Kev 0.8B | 63.1% | 86 ms |
| GLiNER2.5-Decide (English only) | 52.4% | 208 ms |
| Laya (English) | 42.9% | 41 ms |
| Julia-1 | 31.5% | 12 ms (GPU) |
The author's conclusion: if Jev-like accuracy is a requirement, Laya, Julia-1 and the 0.6B Decision 3.0 Edge are not usable out of the box. He also notes that judging these small encoders by the accuracy column alone misreads them: they are built to be fine-tuned and run at the edge. Fine-tuning was not tested.
Arize (October 2026) ran a bracket of hallucination detection on RAGTruth, LLM-evaluator suites and agent routing. Balanced accuracy on RAGTruth was 85.5% for Jev, 85.2% for Clef 27B, 76.8% for Kev 9B and 59.8% for laya-multilingual. Laya was the fastest model at 0.04 s p50 on an M4 Max, and had the highest calibration error (0.330).
systemonemodels.ai lists 29 open Jev alternatives with the accuracy each publisher reports. Those figures come from different test sets, so they cannot be compared with each other.
How to choose
- Speed, CPU or on-device: the encoders. Laya and Julia-1 answer in tens of milliseconds, and Laya has the most ports, from Mac to the browser. Plan to fine-tune or to keep a fallback for hard cases.
- Accuracy without training: the 4B and larger models. They need a GPU or an Apple Silicon Mac with enough memory, and they are several times slower.
- No servers: Jev, Clef on Workers AI, or a hosted Laya API.
Whichever you pick, run a few hundred of your own labelled examples through two or three candidates. The rankings above come from small or synthetic sets, and their authors say so.
More ways to use Laya