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
layaAsRagJudge
vishalmysore
Verifies RAG claims in the browser using Laya decisions and evaluates accuracy against labeled claims
laya-cuda
adamjen
Benchmarks Laya typed-decision checkpoints with a CUDA backend on NVIDIA GPUs
laya-exp
changtimwu
Experiments comparing an MLX Laya port with Jev and a TypeSafe-compatible Laya client
laya-lab
Alevsk
Runs Laya locally with reproducible scenarios that explore its capabilities on Apple Silicon
laya-inference-engine
aaditagrawal
Provides a GPU inference engine and performance modes for Laya on NVIDIA Blackwell
verdict
Manavarya09
Provides calibrated zero-shot and fine-tuned typed decision models as an alternative to Jev
kime
tamnd
A Rust inference engine and HTTP server runs Laya checkpoints across CPU, CUDA, and Apple GPUs
laya-vs-dijkstra
antonellof
Compares Laya MLX pathfinding decisions with Dijkstra on seeded weighted mazes
Laya-Finetune
zamax14
Fine-tunes and evaluates multilingual Laya for support-ticket classification
cbjev
tomek7667
Implements a Jev-compatible decision model and benchmarks it directly against Laya
apm-laya-triage
danielmeppiel
Benchmarks local Laya issue classification against labels in the Microsoft APM corpus
laya-system1-snake
Kamandlou
A terminal Snake game uses Laya MLX to choose moves from structured game-state features
laya-eval
yliuai
Fine-tunes Laya to play Snake on a Mac and measures base vs fine-tuned decision accuracy
chong-laya
cloudwallker
Provides a local interface for inspecting Laya shopping decisions, probabilities, latency, and raw responses
jev_laya_studies
professortavares
Explores structured text decisions, explainability, and comparisons involving Jev and Laya
system-one-web-ai
Josuesp1620
An interactive web guide measuring Laya on Spanish customer-service messages and explaining its implementation
laya-persian-benchmark
alipyth
Benchmarks Laya's Persian support-triage decisions across eight language-specific message families
laya-evals
Gjusev
Evaluates Laya decision models and audits confidence calibration for automated decisions
laya-evals-gate
Gjusev
Compares Laya evaluation metrics against baselines and fails CI when accuracy or calibration regresses
typed-decision-slay-the-spire
MindrLabs
Compares Convai Laya and Nimble models playing Slay the Spire, including fine-tuned runs
google-adk-laya-router
derrickqin
A Google ADK example routes requests among Gemini tiers using Laya and evaluates routing results
laya-jev-benchmark
eugeneproai
Compares local Laya and hosted Jev across decision scenarios with benchmark telemetry
ferrum
SampleBias
Plans a Hermit unikernel whose resource policies are selected by an external Laya process
laya-vs-llms-benchmark
stuaninauts
Compares Laya with machine-learning and language models on financial complaint triage
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