Read & learn
Written guides.
Understand Jev, one idea at a time. Walkthroughs, recipes, and writeups — our notes first, the original next.
A public registry · independent & community-run
Explore the Jev ecosystem.
Learn, build, and play with Jev — from deep technical guides to creative experiments.
Watch & learn
Video guides.
See an idea take shape. Tutorials, demos, and deep dives, organized by topic and credited to their creators.
Make & explore
Experiments.
See what builders made with Jev. Games, tools, repositories, articles, and more — traced to their sources.
The directory03 / 03 · Experiments
Find something worth exploring.
Every category, from games and tools to repositories and writeups, traced to a primary source.
Same ideas.
Different paths.
This is the full registry, not just the featured picks. Figures like speed, cost, and stars are maker-reported or captured snapshots, not JevMade measurements.
1612 experiments · showing 901–960
Jev considers several questions about the board at once to choose Snake's next move.
SOURCE SCREENSHOTEXPAND ↗
from the source
git clone https://github.com/siroccomask/snake-jev.gitcd snake-jevuv synccp .env.example .env
Jev presses the controls of a small virtual creature while deterministic code runs its world.
SOURCE SCREENSHOTEXPAND ↗
Play either side of a 3D chessboard against Jev, or watch two Jev players face off.
SOURCE SCREENSHOTEXPAND ↗
Two Jev agents play chess, choosing every move through a typed Choice question.
SOURCE SCREENSHOTEXPAND ↗
Jev controls a Minecraft Java player through tasks such as gathering lumber and building a Canadian flag.
SOURCE SCREENSHOTEXPAND ↗
from the source
{ "model": "jev-latest", "state": { "scenario": "lumber", "goal": "Find trees, collect 10 new logs, then return to the starting position.",
CartShield assigns a fraud disposition to checkout activity for small online shops.
SOURCE SCREENSHOTEXPAND ↗
Crowdcheck simulates how 10,000 AI readers might react to a post before publication.
SOURCE SCREENSHOTEXPAND ↗
Twelve deliberately unqualified AI councillors vote on unusually specific questions.
SOURCE SCREENSHOTEXPAND ↗
from the source
TYPESAFE_API_KEY=your-key-hereTYPESAFE_MODEL=jev-latest
Harness Judge labels each agent step as acceptable, worth retrying, needing escalation or ready to stop.
SOURCE SCREENSHOTEXPAND ↗
HireSignal gives resumes a first-pass fit assessment and an interview recommendation.
SOURCE SCREENSHOTEXPAND ↗
Human Compiler turns prose into compiler-style diagnostics for traits such as passive aggression and corporate jargon.
SOURCE SCREENSHOTEXPAND ↗
from the source
Compiling input.txt (--mode corporate) Lexing 0.4ms 4 paragraphs, 7 sentences, 62 words Analyzing 812ms jev-1.13.0, 45 questions, 1203 tokensPASSIVE_AGGRESSION ████████░░ 0.82CORPORATE_BULLSHIT █████████░ 0.94
Upload a dataset, pose a question and watch Jev classify every row.
SOURCE SCREENSHOTEXPAND ↗
from the source
CONVEX_URL=https://<deployment>.convex.cloudCONVEX_WRITE_SECRET=<operator-provisioned-secret>CONVEX_DEPLOY_KEY=<Convex production deploy key, Vercel Production only>OPENROUTER_KEY=<operator-provisioned-secret>JEV_API_KEY=<operator-provisioned-secret>
A 3D robot companion reacts through typed Jev decisions rather than generated dialogue.
SOURCE SCREENSHOTEXPAND ↗
An interactive room changes in response to the visitor's words, using presets or live Jev decisions.
SOURCE SCREENSHOTEXPAND ↗
A public demo uses Jev to review an advertisement before it runs.
SOURCE SCREENSHOTEXPAND ↗
An interactive workspace for browsing Jev Board datasets and asking Jev questions about them.
SOURCE SCREENSHOTEXPAND ↗
A first Jev tutorial in TypeScript and Bun, starting with the question of whether a hot dog is a sandwich.
SOURCE SCREENSHOTEXPAND ↗
from the source
# Playground の Noul のサンプル「ホットドックはサンドイッチか?」を API でbun run noul1bun run noul2 # placefolder 使用bun run noul3 # 並列問い合わせ# noul1 を改造して、日本語のテスト「バナナはおやつに入りますか?」
A Dart collection tests Jev across several small demonstrations.
SOURCE SCREENSHOTEXPAND ↗
A single screen compares Jev and an LLM on emotion shifts and response speed for the same statement.
SOURCE SCREENSHOTEXPAND ↗
Jev Paper Judge gives rapid feedback on an academic paper.
SOURCE SCREENSHOTEXPAND ↗
Jev chooses a musical plan from fixed options, then code renders the score, audio and MIDI.
SOURCE SCREENSHOTEXPAND ↗
from the source
“Philip Glass” src/ │ style string ▼ ┌─────────────────────────────┐ Planner interface planner/Planner.ts │ JevPlanner │ Heuristic- │ plan(input) → CompositionPlan planner/JevPlanner.ts
A job-application demo that asks Jev whether a role is worth applying for.
SOURCE SCREENSHOTEXPAND ↗
A public demo turns Jev into a jury for user-related decisions.
SOURCE SCREENSHOTEXPAND ↗
Build custom Choice, Score and Noul questions and watch their probability distributions update live.
SOURCE SCREENSHOTEXPAND ↗
A CLI and API assess an occupation's exposure to AI layoffs, future path, human accountability and resilience.
SOURCE SCREENSHOTEXPAND ↗
LaneBreak assigns support tickets a priority and destination.
SOURCE SCREENSHOTEXPAND ↗
MCPMatch uses a two-stage process to pair a user's goal with tools from an MCP catalog.
SOURCE SCREENSHOTEXPAND ↗
Live Bitcoin, Ether and XRP prices feed a Jev demonstration that chooses between buying and waiting.
SOURCE SCREENSHOTEXPAND ↗
A clinic-triage demo that asks Jev which cases need attention.
SOURCE SCREENSHOTEXPAND ↗
from the source
npm i && npm run dev# optional: echo KEY > .env.local as TYPESAFE_API_KEY=...
A conversational search prototype helps readers find relevant material among hundreds of blog posts and articles.
SOURCE SCREENSHOTEXPAND ↗
Spendbrake watches an agent's budget and chooses whether to continue, downgrade the model or stop.
SOURCE SCREENSHOTEXPAND ↗
Swarmrouter sends each task to a research, coding, browser, support or writing agent.
SOURCE SCREENSHOTEXPAND ↗
TrustGate screens trust-and-safety decisions for independent media publishers.
SOURCE SCREENSHOTEXPAND ↗
TypeSafe Typewriter reruns sixteen judgments about a piece of text after every keystroke.
SOURCE SCREENSHOTEXPAND ↗
A classifier creates pixel art by answering 256 pixel questions in parallel and refining the result over several passes.
SOURCE SCREENSHOTEXPAND ↗
Ask a question and receive Yes, No or Maybe, with live web search available for current facts.
SOURCE SCREENSHOTEXPAND ↗
This experiment tests Jev on financial-market prediction and reports that its forecasts perform poorly.
SOURCE SCREENSHOTEXPAND ↗
jev-alpha-bench separates news comprehension from market recall to test whether Jev can predict stock returns.
SOURCE SCREENSHOTEXPAND ↗
This benchmark compares Jev, Gemini, and GPT on structured annotation of São Paulo court judgments.
SOURCE SCREENSHOTEXPAND ↗
from the source
uv syncuv run python scripts/01_coletar.pyuv run python scripts/02_anotar.py jev-latest jev-preview gemini-3.8-flash gpt-5.6-luna... ref-gpt-5.6-sol ref-gemini-3.1-prouv run python scripts/03_referencia.py # lista divergências e auditoria
This project experiments with Jev as an evaluator rather than a task-solving model.
SOURCE SCREENSHOTEXPAND ↗
A position paper arguing that Jev-style models need fuzzy and Hidden Markov primitives before they settle on crisp decisions.
SOURCE SCREENSHOTEXPAND ↗
from the source
git clone https://github.com/dnakhoa/jev-deferred-crispificationcd jev-deferred-crispificationpython3 experiments/run_all.py # E1–E5 + lemma checks, CPU, ~2 min
Jev picked the winner in 64.5% of 10,984 randomized Upworthy headline tests.
SOURCE SCREENSHOTEXPAND ↗
This project tests whether Jev can resolve ambiguous Japanese addresses against Japan Post's KEN_ALL data.
SOURCE SCREENSHOTEXPAND ↗
from the source
npm installscripts/download-data.sh # data/utf_ken_all.csv, data/jigyosyo_utf8.csv を取得...export TYPESAFE_AI_API_KEY=... # Jev を使う場合のみnpm run build # dist/cli.js
jev-lab collects TypeScript experiments on Jev's behavior, accuracy, and response latency.
SOURCE SCREENSHOTEXPAND ↗
A phishing test that pits Jev against Claude Haiku 4.5 on 2,000 emails.
SOURCE SCREENSHOTEXPAND ↗
from the source
cp .env.example .env # then fill in the keys (baseline used: claude-haiku-4-5...uv run prepare_data.py # download + checksum + data/emails.jsonluv run net_floor.py # network round-trip floor to each API hostuv run run_jev.py --limit 10 # smoke test, prints raw answersuv run run_jev.py # pass 1, all 2 000 emails
This reproducible MuJoCo pilot compares Jev, Claude Haiku, and reactive rules on pick-and-place control.
ACTUAL RECORDING12 SEC ↗
from the source
.venv\Scripts\python.exe -m zipfile -e results\evaluation_traces.zip outputs\comparison.venv\Scripts\python.exe demo.py render outputs\comparison\eval_jev_ordinary_1001.jsonl
This game playground compares Jev with other evaluators using explicit states, legal moves, and observable outcomes.
SOURCE SCREENSHOTEXPAND ↗
from the source
flowchart LR state[Current state] --> actions[Legal actions] actions --> model[Model evaluation] model --> choice[Validated choice] choice --> next[Next state]
HackSing puts Jev through 50 tests and publishes the findings in a 52-page Chinese-language report.
SOURCE SCREENSHOTEXPAND ↗
This experiment measures Jev as a lower-cost model router on the RouterArena benchmark.
SOURCE SCREENSHOTEXPAND ↗
Can Jev tell a real credential from an innocent string? This benchmark tests it on file snippets.
SOURCE SCREENSHOTEXPAND ↗
from the source
echo "TYPESAFE_API_KEY=..." > .envuv run python main.py
A test of whether Jev can tell genuine shadcn-ui lint problems from false alarms.
SOURCE SCREENSHOTEXPAND ↗
from the source
node extract-design-system.mjs --repo <lint checkout> # only when the lint repo change...export TYPESAFE_API_KEY=... # never commit itnode run-rule-cases.mjs rule-cases.jsonl --repo <lint checkout> [--limit n]
This notebook compares zero-shot spam judgments from Jev with conventional TF-IDF classifiers.
SOURCE SCREENSHOTEXPAND ↗
from the source
# New output directory: a three-message smoke test, then the full comparison.JEV_CONTEXT_OUTPUT_DIR=experiments/jev-context/rerun \ uv run python experiments/jev-context/evaluate.py --limit 3JEV_CONTEXT_OUTPUT_DIR=experiments/jev-context/rerun \ uv run python experiments/jev-context/evaluate.py
This classifier distinguishes agent-written from human-written pages in the collusion.wiki corpus.
SOURCE SCREENSHOTEXPAND ↗
from the source
ground_truth.py # the per-page agent/human label join (the only "ground truth" logic)baseline_regex.py # content-only floor: regex tells -> shows text alone is blindjev_client.py # minimal TypeSafe Jev client (noul primitive, plain HTTP)qwen_client.py # local Qwen3.8-Flash-Next client (OpenAI-compatible)compare.py # the head-to-head harness + metrics table + JSONL dump
This repository tracks ongoing research into Jev and System One models as a set of Markdown slides.
SOURCE SCREENSHOTEXPAND ↗
This app extracts data from a document according to a supplied JSON schema using parallel constrained decoding.
SOURCE SCREENSHOTEXPAND ↗
A Jev-inspired experiment that makes visual decisions on an iPhone using Qwen3-VL, not TypeSafe's service.
SOURCE SCREENSHOTEXPAND ↗
This study tests Jev as a fast gate for suspicious agent actions in SHADE-Arena.
SOURCE SCREENSHOTEXPAND ↗
from the source
conda create -n shade-arena python=3.11conda activate shade-arenapip install -r requirements-jev.txt # minimal pinned deps; upstream requirements.t...conda env config vars set PYTHONUTF8=1 # task data is UTF-8; Windows defaults to cp125...conda deactivate && conda activate shade-arena
This Rust port runs Jev-style Choice, Score, and Noul evaluations through ordinary language models.
SOURCE SCREENSHOTEXPAND ↗
from the source
use system_one_adapter::{ Noul, ProviderName, SystemOneAdapterClient, SystemOneArgs,};let client = SystemOneAdapterClient::new(true, system_one_adapter::AnswerMode::Probabili... .normalize_probabilities(true);
system-one-gemma adds a scoring head to Gemma 3 270M for calibrated decisions without text generation.
SOURCE SCREENSHOTEXPAND ↗
from the source
from infer import load_trained_model, scoretok, model = load_trained_model("./pretrained-scorer")probs = score(model, tok, state="Patient has chest pain radiating to left arm, ST elevation on ECG", question="What is the triage level?",
These charts place Jev's Thai standardized-exam results alongside those of 110 other models.
SOURCE SCREENSHOTEXPAND ↗
Jev, briefly
A decision model, not a chatbot.
A chat model generates text token by token and hopes you parse it. Jev never generates a word. Your code sends the state of the world plus typed questions; Jev returns every answer in one parallel pass — typed values with calibrated probabilities and a confidence score your code can trust. About 70–500 ms end to end, $0.042 per million input tokens, output free. Trained with what TypeSafe calls Reinforcement Learning for Calibrated Decisions.
choice Pick one of these.
You define the options. Jev returns the chosen option, a probability for every option, and a confidence. The workhorse of routing, agents, and games — the answer space is always legal, so there is nothing to parse and nothing to hallucinate.
score Rate this on a rubric.
You describe ordered levels — say trivial / normal / critical. Jev returns the level, the probability of each, and a confidence. Scores turn fuzzy judgment (“how severe is this log line?”) into a number-free decision your code can branch on.
noul Is this statement true?
You assert a statement; Jev returns the probability that it is true — a noul. Moderation, verification, guardrails, “does this diff actually fix the bug?”: one question, one calibrated probability.
Official material lives at typesafe.ai and docs.typesafe.ai. JevMade is an independent community registry — not affiliated with or endorsed by TypeSafe AI.
Recurring lessons
Patterns that keep showing up.
One decision per tick
Games, drones, trading bots, and browser agents all converge on the same loop: serialize the state, ask one decisive question, act, repeat. Jev’s latency makes the loop feel instant — the model lives inside the control loop, not outside it.
Confidence as a gate
The answer says what; the confidence says whether to act. The most reliable entries threshold on confidence to route edge cases to a slower model or a human — automation with an honest escape hatch.
Shrink the choice space in code
Makers rarely ask Jev to pick from everything. Local tactics prune 225 gomoku moves to ~40; DOM filters turn a page into an element table; code narrows options, Jev judges within them.
Jev judges, LLMs talk
A recurring split: Jev makes every decision cheaply and instantly, and a small LLM is only invoked when a human-facing string must actually be written. Decision and generation are separate budgets.
Made something with Jev?
Send the link — repo, live demo, post, or video — plus a line on what it does and which primitives it uses. Every entry is verified against its primary source before it ships.
Submit your experiment → hello@JevMade.com