What is the JevAny project?

Open infrastructure for training, evaluating and deploying System 1 decision models across language and multimodal backbones, from the SimpleJev project, covering data preparation, model adaptation and evaluation.

How does JevAny relate to Jev?

It is infrastructure for the same System 1 idea: instead of generating text, a model takes a state, a question and candidate options and directly returns a choice with probabilities, which the README frames as decision-model behaviour.

What can I build with JevAny?

The README lists routing support tickets, selecting tools and choosing a robot's next action, using either a released checkpoint or training on your own data.

What does the JevAny API look like?

Per the README, one API takes the state, question and candidate options and returns both a choice and its probabilities. The full API reference is linked at simplejev.org/JevAny/docs/API.html.

How do I run JevAny locally?

The README has Quickstart and Run locally sections and targets Python 3.12+, but the digest excerpt is truncated before those commands, so follow the repository README or the website for current steps.

Where are JevAny's results reported?

The repo links checkpoints in a Hugging Face collection and bar charts comparing checkpoints and baselines on Transfer and JevBench, with interactive benchmark results on the simplejev.org homepage.

What licence and activity level does JevAny have?

Apache-2.0 with Python as the language, 104 stars and 11 forks, created 2026-09-22 and last pushed 2026-10-16, so the project is active and maintained by the SimpleJev organisation.

Jev directory

CommunityJev Repos & SDKs

JevAny — SimpleJev

Open infrastructure from the SimpleJev project for training, evaluating and deploying System 1 decision models across language and multimodal backbones: one API takes a state, question and candidate options and returns a choice with probabilities.

See highlights Open github.com

Stars: 105
Language: Python
License: APACHE-2.0
Reading time: 2 min
Published by: Community
Author: SimpleJev
Last verified: Oct 2026

SimpleJev/JevAny

Organization repository on GitHub

View on GitHub

Open infrastructure for training, evaluating, and deploying System 1 decision models across language and multimodal backbones.

Forks: 11
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

Tagscommunity github python system-one lora

Works forJev for Routing Jev for Tool calling Jev for Realtime control

Used inJev for Software & DevTools

Languages: Python
Platforms: API
Capabilities: Open sourceRuns locally

Highlights

Covers the full loop for System 1 decision models: data preparation, model adaptation and evaluation.
One API takes the state, question and candidate options and directly returns a choice and its probabilities.
Use a released checkpoint or train on your own data to route support tickets, select tools, or choose a robot's next action.
Checkpoints are published in a Hugging Face collection and compared against baselines on Transfer and JevBench.
Targets Python 3.12+ under Apache-2.0, with a homepage, API docs and an examples gallery at simplejev.org.

Watch out

README reviewed 2026-10-17; the excerpt is truncated during the results section and before the Quickstart and Run locally commands, so install steps, training costs and benchmark numbers are not in the digest and all claims are the project's own.

Questions about this entry

Short answers drawn from the linked project or page.

Other listings of this project

Forks and copies of this project in the directory. This entry is the canonical one.

JevAny

105

Open training and deployment stack for System 1 decision models: one state-plus-question-plus-options API that returns a choice with probabilities for tickets, tools, or robot actions.

Jev Demos & Experiments#community#jevbest#catalog-import

Community

weitianxin/JevAnyPython

From the Jev guides

Original write-ups on what this entry is for.

The Open Reimplementations: Local Models That Speak System One

Sep 22, 2026

Closed weights have not stopped the ecosystem from rebuilding the shape in the open. API-compatible servers, retrained heads on small models, and inference-time tricks that skip generation entirely.

Jev Repos & SDKs#open-models#research#local

Read article

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai