visual-jev

Where it calls Jev

@app.post("/v1/systemone")

src/visual_jev/api.py:576

The link points at the commit we read, so the line number still holds.

What it asks Jev

These question sets are lifted from this project's source exactly as written.

import { experimental_evaluate as evaluate } from 'ai';

const { answers } = await evaluate({
  model: 'typesafe-ai/jev',
  state,
  questions: {
    composition: { type: 'choice', instructions: 'Is the composition balanced?',
      criteria: { balanced: 'The visual weight is balanced', left-heavy: 'Weight sits mainly on the left', unclear: 'The image cannot be assessed' } },
    legibility: { type: 'boolean', instructions: 'Is the title legible?' },
  },
});

Other projects in this category

• NanoJev ▶ — A nano replica of Jev: parallel decisions, dynamic candidates, and an end-to-end training pipeline.
• kev ▶ — tiny Jev-like model built on top of Qwen2.5-0.5B you can train and run on your MacBook
• reflex ▶ — A small open decision model: state + typed questions -> calibrated probabilities. A Jev / System One re-creation on Qwen3.5.
• von ▶ — The open-source System One decision model. Sub-15ms, non-autoregressive, local drop-in alternative to TypeSafe Jev.
• OpenDecision ▶ — OpenDecision is an open-source semantic decision engine like typesafe's jev.
• jev_local ▶ — Replicating Jev with a local LLM
• OpenJev ▶ — Jev-compatible System 开源Jev
• diffusiongemma-jev-macos ▶ — Local JEV-style decisions with DiffusionGemma on Apple Silicon, with benchmarks and coding-agent examples.