EVERY ENTRY HAS RECEIPTS

AWESOME
JEV

Open-source directory

Every project here really calls Jev in its source · refreshed daily

788projects
182runnable here
765code receipts

ALL 788 PROJECTSEvery project

scripts/jev_probe.mjs
questions: {
  requested: boolean,
  issued: boolean,
  route: choice,
  // +1
}
runnable
@TianyuCodings1.1k

NanoJev

A nano replica of Jev: parallel decisions, dynamic candidates, and an end-to-end training pipeline.

Open reproductionsPython
playground/src/lib/kev.ts
questions: {
  department: choice,
  return_reason: choice,
  requested_resolution: choice,
  // +2
}
runnable
@jaredpalmer617

kev

tiny Jev-like model built on top of Qwen2.5-0.5B you can train and run on your MacBook

Open reproductionsPython
docs/app.js
questions: {
  queue: choice,
  escalate: boolean,
  urgency: score,
  // +1
}
runnable
@kshetrajna1280

reflex

A small open decision model: state + typed questions -> calibrated probabilities. A Jev / System One re-creation on Qwen3.5.

Open reproductionsPython
examples/intent_routing.py
questions: {
  intent: choice,
  needs_reasoning: score,
}
runnable
@wfzyx48

von

The open-source System One decision model. Sub-15ms, non-autoregressive, local drop-in alternative to TypeSafe Jev.

Open reproductionsPython
examples/m3_typesafe_sdk_demo.py
questions: {
  department: choice,
  urgent: boolean,
  frustration: score,
}
runnable
@deepanwadhwa29

OpenDecision

OpenDecision is an open-source semantic decision engine like typesafe's jev.

Open reproductionsPython
tools/verify_api.py
questions: {
  binary: boolean,
  rating: score,
}
runnable
@Argos111115

jev_local

Replicating Jev with a local LLM

Open reproductionsPython
scripts/test_official_jev.py
questions: {
  department: choice,
  frustration: score,
  is_urgent: boolean,
}
runnable
@GPT-AGI2

OpenJev

Jev-compatible System 开源Jev

Open reproductionsPython
src/diffusion_jev/benchmark_cases.py
questions: {
  enabled: boolean,
  color: choice,
  progress: score,
}
runnable
@Saik0s2

diffusiongemma-jev-macos

Local JEV-style decisions with DiffusionGemma on Apple Silicon, with benchmarks and coding-agent examples.

Open reproductionsPython
tests/test_core.py
questions: {
  composition: choice,
  legibility: boolean,
}
runnable
@andrueandersoncs1

visual-jev

Image-native typed decisions with shared visual encoding and Qwen3-VL

Open reproductionsPython
scripts/jev_conformance.py
questions: {
  billing: boolean,
  urgent: boolean,
  team: choice,
  // +1
}
runnable
@zerodegress1

jevinf

Jev-like model inference engine + Jev-compatible API

Open reproductionsPython
benchmarks/build_every.py
"adaptation": "Native Noul replaced with binary Yes/No Choice; state/question unchanged",
@TheoLeeCJ2.0k

SemIf

Semantic ifs from open models, on a 3090 at home. Independent; not affiliated with Jev or TypeSafe.

Open reproductionsPython
openjev/api.py
"""Jev-compatible HTTP API: POST /v1/systemone and GET /v1/models.
@razorback16146

openjev

Open, Jev-compatible System One decision server on DiffusionGemma

Open reproductionsPython
src/jeff/server/app.py
@app.post("/v1/systemone", response_model=SystemOneResponse)
@logan-markewich125

jeff

A self-hosted drop-in replacement for TypeSafe's jev, powered by GliFormer.

Open reproductionsPython
scripts/exp_external_eval.py
"jev": j_val,
@Heman10x-NGU80

openJev-verdict-2.0

Calibrated 151M Non-Autoregressive Decision Engine beating TypeSafe Jev & Laya on LocalLLaMA/typed-decisions (77.10% acc, 0.0636 Brier, 0.0144 ECE)

Open reproductionsPython
openjev/cli.py
v = sub.add_parser("serve", help="HTTP server with the model loaded once (POST /score, /v1/systemone)")
@daseinlabs58

open-jev

Open Jev implementation with custom finetuning

Open reproductionsPython
scripts/exp_external_eval.py
"jev": j_val,
@Heman10x-NGU28

Verdict-open-jev

Non-autoregressive decision engine on ModernBERT (151M) with calibrated uncertainty (RLCD), TypeSafe AI Jev benchmark audit, and in-browser WebGPU playground

Open reproductionsPython
pyproject.toml
keywords = ["jev", "typesafe-jev", "open-source-jev", "jev-alternative", "system-one", "system-one-model", "llm", "classification", "structured-decisions", "typ
@ikermoel16

open-alternative-jev

Open-source alternative to TypeSafe's Jev: a System One style model layer that gives typed, calibrated decisions from any open-weights LLM in one forward pass (HF + vLLM), with honest benchmarks

Open reproductionsPython
make_report.py
for k, name in [("noul", "Noul (yes/no)"), ("score", "Score (rubric level)"), ("choice", "Choice")]:
@mithalouni13

system-one-open

Open replica of TypeSafe's Jev: typed calibrated decisions in one forward pass, on Gemma 4 E2B / Gemma 3 270M (Modal)

Open reproductionsPython
openvons/lm/api/server.py
# ---------------------------------------------------------------- TypeSafe Jev 互換 (POST /v1/systemone)
@genai-craft9

openvons

openvons (open-Jev): 有限選択肢に確率で答える判断層 — テキスト / 画像 / 日本語音声コマンド

Open reproductionsPython
scripts/race-omp-backends.py
TYPESAFE = os.environ.get("TYPESAFE_BASE_URL", "https://api.typesafe.ai").rstrip("/")
@kvnloo3

z0intelligence

Can we run something like Jev on a 3090 at home?

Open reproductionsPython
src/jevify/__init__.py
from .schema import Choice, Noul, Question, Score, question_from_dict
@Mintzs3

jevify

An optimized inference engine to turn LLMs into Jev-like machines: optimized for quick, lightweight, and accurate decision-making, classification, and scoring

Open reproductionsPython
app/main.py
"app": "jev",
@sabeel1113

OpenSourceJev

Turning an LLM model into a Jev like System.

Open reproductionsPython
server/__main__.py
if self.path != "/v1/system_one":
@xingwudao3

OpenJev

OpenJev: an independent Jev-inspired System One decision API based on TypeSafe.ai concepts. Choice, score and noul primitives, local mock server, Python and TypeScript SDKs. Real inference planned; not affiliated with TypeSafe AI.

Open reproductionsPython
system_one.py
"""System One / "Jev"-style scorer.
@akash-kamat2

system-one-gemma

Open-source Jev-style System One decision model. Gemma 3 270M with a scoring head — fast, calibrated decisions in a single forward pass. No text generation. Inspired by TypeSafe.ai's Jev.

Open reproductionsPython
benchmarks/build_every.py
"adaptation": "Native Noul replaced with binary Yes/No Choice; state/question unchanged",
@leesk2122

JEV-CPU

Run SemIf (Jev-style semantic-if decisions) on a CPU — no GPU. Reads typed option probabilities straight from an open model in one forward pass, plus a web UI.

Open reproductionsPython
jev_llm.py
Requires TYPESAFE_API_KEY and: pip install typesafe-sdk
@afanjul1

jev-llm

Fake autoregressive language model powered by TypeSafe Jev

Open reproductionsPython
jev.py
Endpoint: POST https://openrouter.ai/api/alpha/decisions (model ~typesafe/jev-latest).
@burgerwdev1

what-is-jev

Jev notes: fourteen decisions I ran through it

Open reproductionsPython
src/index.ts
import { TypeSafeClient, TypeSafeDecisionModel } from "@effect/ai-typesafe";
@ChristianAlexander1

effect-jev-cwe

A demonstration of the Jev System 1 model in Effect, matching vulnerabilities to their underlying CWEs

Open reproductionsTypeScript
src/decider.rs
"jev" => Some(Self::Jev),
@corasan1

arjev

Argent + Jev

Open reproductionsRust
openjev/audit.py
j = row.get("jev")
@DECRUX98121

openjev-lm

open-Jev LM arm: Qwen2.5-0.5B + LoRA reproducing a hosted decision model's judgment at 92.9% on hand-labelled gold - trained overnight on a 6-vCPU CPU-only host, $0/call. Paper, corpora, harnesses, receipts.

Open reproductionsPython
main.py
from typesafe_sdk import Choice, TypeSafeClient
@erik-dunteman1

ChatJev

Turning the Jev classifier model into an autoregressive next token predictor

Open reproductionsPython
src/systemone_lite/api.py
"""FastAPI server exposing Jev-compatible POST /v1/systemone."""
@fritzprix1

systemone-lite

Toy local System One–style decision API (Jev-shaped). Not affiliated with TypeSafe.

Open reproductionsPython
bench_bridge.py
"http://127.0.0.1:8768/v1/systemone",
@rorshopping1

jev-browser-local

Run jev-browser on a fully local JEV-style decision engine (no cloud API). Warm-browser fork, VRAM guard, measured benchmarks, run traces.

Open reproductionsPython
foq/__init__.py
from .engine import FoqEngine, FoqConnectionError, SystemOneResponse
@yohanargentina-oss1

Foq

⚡ Foq — the FREE, local, open-source alternative to Jev. Typed System 1 decisions in ~25 ms — no waitlist, no cloud, no per-token cost. foq.fr

Open reproductionsPython
jevforge/serve.py
POST /v1/systemone   {state, model, questions} -> {model, answers, usage}
@zwliJay1

jev-forge

An open training and inference stack for Jev-style decision models. Train models to score dynamic candidate branches from a shared prefix, with support for high-cardinality choice, calibration, and fast batched inference.

Open reproductionsPython