Jev Decision Index
Hugging Face Space indexing open reproductions of Jev and comparing them on a versioned decision benchmark with calibration metrics.
Overview
Jev Decision Index is a static Hugging Face Space that tracks open reproductions of TypeSafe's Jev and benchmarks them. Its metadata lists dozens of models, including GLiNER2.5 variants, Laya, Intern-Decision, Decision-1.0, decider, pplx-decider-v1-27b, kev, jpt, RLCD fine-tunes and base Qwen and Gemma models. A list summary describes it as tracking 70 open reproductions as of version 0.2.1 and serving as the reference list for the replication wave.
It compares Jev with open decision models on a versioned benchmark suite with calibration metrics and documented methods. One summary cautions that open-model inference timings and Jev's hosted HTTPS latency are not directly comparable. It is useful for anyone choosing an open alternative or following the field.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Other | How do open Jev-style models compare with Jev on accuracy and calibration?(inferred) | — |
How it works
- Collect open decision models that reproduce Jev's interface
- Run each on a versioned benchmark suite
- Report accuracy and calibration metrics with documented methods
- Publish results and news in the Space
Strengths
- Single reference list of open reproductions
- Calibration metrics, not just accuracy
- Versioned benchmark
Caveats
- Latency of local open models and hosted Jev is not directly comparable
- Content changes as new models are added
Getting started
Open https://huggingface.co/spaces/multimodalart/jev-decision-index in a browser.At a glance
- Type
- Article
- Status
- production
- Runtime
- Static Hugging Face Space
Use cases and integrations
How curated lists describe it
Index: an independent Space tracking 70 open reproductions of TypeSafe Jev as of version 0.2.1, the reference list for the replication wave.
yibie/awesome-jevHugging Face Space comparing Jev with open decision models on a versioned benchmark suite, with calibration metrics and documented methods; open-model inference timings and Jev's hosted HTTPS latency are not directly comparable
AnotiaWang/awesome-decision-models
Frequently asked questions
What does the index track?
Open models that reproduce Jev's typed-decision interface, with benchmark and calibration results.
Can I compare latency?
Only with care; local inference timings and hosted API latency are measured differently.
Do I need anything to use it?
No, it is a public web page on Hugging Face.
Related projects
Evaluation suite for decision models with 23,900 pinned rows across nine task families and three primitives, plus a leaderboard.
ProjectbetaPythonJev-compatible open System One engine with a Claude Code-style terminal REPL showing live probabilities, plus an experiment harness.
ProjectprototypePythonLlamaIndex notebook comparing Jev with open decoders, encoders and specialised tools on document-pipeline Choice tasks.
ProjectresearchPythonBenchmark harness comparing Jev with OpenRouter models and open Jev-shaped models on accuracy, calibration, latency and cost.
ProjectresearchPythonReproducible measurements of Jev's real cost, latency, response shapes and accuracy, plus 13 decision models on the same 633 items.
ServiceresearchShort report comparing Jev and many other fast decision classifiers with an LLM on grading section relevance in technical documents.
Projectresearch