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.
472 projects
laya-jev
pcbrom
Teaches typed decision models through a comparison of Laya and Jev, including architecture, calibration, and fine-tuning
laya-test
supermy
Runs quantized Laya inference locally on Android and Termux using native ONNX Runtime
Jev_vs_Laya_decision_games
minanayernia
Game-theory experiments compare Jev and Laya across six games with analysis and a browser app
system-one-triage-eval
Pedro-Yataco
Evaluates ticket-triage performance across Laya, Jev, and language-model backends
laya-coreml-experiment
kazuki-oshino
A SwiftUI iOS app compares on-device Core ML inference for Laya and Kev
Jev-AI-vs-Laya-AI
basavaraj1997
A comparison guide and local evaluation suite examines Jev and Laya decision engines
laya-hate-speech-benchmark
jasontiw
A reproducible benchmark compares Laya hate-speech detection with other classifiers
laya-test
psenger
Tests Laya decisions, confidence calibration and latency on Apple Silicon GPUs
laya-ggml
liujp
Implements a C inference engine and HTTP service for the Laya decision model
laya-stack
spaxfiz
Provides an MLX inference studio, evaluation benchmarks, and data workflows for Laya decision models
laya-01
D-Ajay-Kumar
Benchmarks and fine-tunes Laya-format decision models for column relevance and yes/no tasks
laya-local-decisions
karti-ai
Documents local Laya deployment, API checks, benchmarks, and evaluation on CPU and NVIDIA GPUs
laya-snake-benchmark
IYIcode
Benchmarks Laya on Snake gameplay with recorded results and browser-based reruns
Benchmark-laya-ml-Autoregressif
Takou07
Benchmarks Laya against a classic machine-learning model and an autoregressive model
laya-f1-strategy-pOc
MiguelMF-Coder
Replays historical Formula 1 races to study typed strategy decisions, backtests, and calibration
sys1-eval
David-DL-Space
Compares Laya and Jev on typed decisions for LLM agent harnesses
S1Q
CYMCharming
Implements and evaluates low-bit quantization for typed decision models including Laya
jev-smart-home-lab
SafalFrom2050
A smart-home simulation that compares Jev API and local Laya decisions
laya-blackjack
hugging-apps
An interactive blackjack demo using a Laya fine-tune to predict action preferences and game outcomes
jev-laya
diptihumraskar
A zero-shot topic classification benchmark comparing Jev, Laya and CLM
layastudio
Lu-cong-1
Provides a local visual workbench for testing, evaluating, and serving Laya decision models
laya-calibration
BargLabs
Publishes a preregistered calibration experiment evaluating Laya as a judge of agent completion reports
laya-qbert
mateusbsmg
Fine-tunes Laya to play Q*bert and reports its performance against a teacher and baseline
LAYA-Decision-Head-
satviksangamkar11
Studies a local decision head and benchmarks it against Laya as a baseline
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