Organization repository on GitHub
OpenSource Jev
- Forks
- 41
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.Homepage
CommunityJev Models & Reimplementations
OpenJev — SiliconLabAI
Express and Vite playground for System One-style decisions, taking a state plus typed questions and returning structured answers. Three swappable backends: per-option LLM micro-scorers, one structured JSON call, or a locally served decider model.
- Stars
- 174
- Language
- TypeScript
- License
- MIT
- Published by
- Community
- Author
- SiliconLabAI
- Last verified
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is OpenJev — razorback16.
About this project
Open source System One–style decision playground: state + typed questions → structured answers. Open — mode parallel or oneshot. decider is an open System One model family (Qwen3.5 fine-tunes). decider-2b v10 includes calibration-aware RL. It speaks TypeSafe’s wire format: POST /v1/systemone. In the UI: set mode = decider, decider url =. Same question types as TypeSafe Jev: choice, score, noul.
Quoted from the project's README. Read the full README.
Works forJev for Classification
Used inJev for Software & DevTools
- Languages
- TypeScript
Highlights
- Backend choice is a UI toggle: parallel per-option micro-scorers, a single structured JSON call, or the Mapika decider model.
- The decider path speaks TypeSafe's /v1/systemone shape, so an existing typed-decision request can be replayed unchanged.
- The same three question types as Jev — choice, score and noul — come back with probability distributions attached.
- Small enough to read end to end: one Express server plus a single evaluate module holding all three modes.
Watch out
The LLM backends only approximate the decider path. Parallel mode scores each option independently and softmaxes the results, oneshot asks one model for the whole JSON structure, and neither is a System One model. The decider backend needs a CUDA GPU with roughly 4 GB free for the 2B checkpoint. The README does not document how the three backends are calibrated against each other, so treat cross-backend comparisons as unvalidated. Copy reviewed against the repository README on 2026-10-04, not from running it.
Questions about this entry
Short answers drawn from the linked project or page.
Parallel mode issues one tiny scoring call per option and softmaxes the results, which keeps each call small when you have many options. Oneshot mode sends a single structured JSON request to any OpenAI-compatible model instead. The README does not compare their agreement or accuracy, so neither can be assumed better.
Yes. The README states the model server needs a CUDA GPU with about 4 GB for decider-2b, installed from the Mapika decider repository with its serve extra and launched through a serve script. Without that hardware, parallel and oneshot still work against any OpenAI-compatible endpoint.
For the decider backend, yes. It speaks TypeSafe's wire format on POST /v1/systemone, and OpenJev forwards to it through POST /api/evaluate with a decider_url, or picks it up from an environment variable. The same choice, score and noul question types are used, though the README does not confirm field names are identical across backends.
No accuracy numbers, evaluation suite, latency figures or test fixtures appear in the README. There is also no license or contribution section in the captured text, and the only walkthrough is a tutorial video rather than written steps, so judging quality means reading the code.
cd OpenJev, run npm install, copy .env.example to .env and set OPENAI_API_KEY, then npm run dev. Open http://localhost:3001 and choose mode parallel or oneshot.
Decider is Mapika/decider, an open System One model family of Qwen3.5 fine-tunes; decider-2b v10 uses calibration-aware RL and speaks TypeSafe's wire format at POST /v1/systemone. Serving it needs a CUDA GPU with roughly 4 GB for the 2B model, started by scripts/serve.sh Mapika/decider-2b 8000.
OpenJev reads OPENAI_API_KEY, optional OPENAI_BASE_URL and OPENJEV_MODEL (default gpt-4o-mini) for the LLM backends, and DECIDER_BASE_URL (default http://localhost:8000) with optional DECIDER_API_KEY for the decider backend.
More Jev resources like this
1.7kGitHub stars
Open-weight 4B and 9B models from TokenRhythm: NeoHorse-1 for tool use, coding, and instruction following, plus NeoHorse-Jev-4B for prefill-only Choice, Noul, and Score decisions.
Jev Models & Reimplementations#community#jevbest
Community
TokenRhythm/NeoHorsePython
TokenRhythm/NeoHorsePython1.5kGitHub stars
A 0.8B decision model that answers first and escalates only when unsure, plus nine LoRA adapters of roughly 41 MB each for guard, triage, tool choice, spam, and similar classification work.
Jev Models & Reimplementations#community#jevbest
Community
firelex/jeffPython
firelex/jeffPython823GitHub stars
A local Jev-compatible POST /v1/systemone server in TypeScript that maps typed questions onto oMLX classification prompts and normalizes logprob-style answers.
Jev Models & Reimplementations#community#typescript#local
Community
githubnext/localjevTypeScript
githubnext/localjevTypeScript2kGitHub stars
DeepOpen is a non-autoregressive decision engine built on Laya that classifies into developer-defined types in one forward pass, with a language router picking between English, multilingual, and typed-decision checkpoints.
Jev Models & Reimplementations#community#jevbest
Community
deepopen-com/deepopenPython
deepopen-com/deepopenPython715GitHub stars
Decision models that answer several choice, score, and noul questions over a state with probabilities in one forward pass, with training, two inference backends, calibration tooling, and the benchmark suite in the repo.
Jev Models & Reimplementations#community#jevbest
Community
InternLM/Intern-DecisionPython
InternLM/Intern-DecisionPython402GitHub stars
Decision models that return typed probabilities for candidate labels from a state, instructions, and options, shipped as LoRA adapters plus a trained decision head over pinned Qwen base weights.
Jev Models & Reimplementations#community#jevbest
Community
Zefan-Cai/Open-JevPython
Zefan-Cai/Open-JevPythonFrom the Jev guides
Original write-ups on what this entry is for.
Jev takes text only; Clef-omni adds image, audio, and video input and Liquid d1 adds images, each keeping the same typed-answer shape. Where that matters, where it does not, and the costs the vendors underplay.
Jev Models & Reimplementations#multimodal#open-models#decision-model
Read article
One model in September, a dozen by October. Clef, Liquid d1, Perplexity, and OpenAI all ship the same typed-answer shape now. How to read the rival benchmarks and pick on your own task.
Jev Models & Reimplementations#decision-model#benchmarks#pricing
Read article
Hundreds of repositories carry the system-one tag, but the ecosystem is a dozen ideas in different sizes. A tour of the families, the licenses, and what a download does not include.
Jev Models & Reimplementations#open-models#hugging-face#local
Read article
From the Jev community
Posts from builders shipping with Jev right now.