Read & learn
Written guides.
Understand Jev, one idea at a time. Walkthroughs, recipes, and writeups — our notes first, the original next.
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.Experiments
Every category, from games and tools to repositories and writeups, traced to a primary source.
Real-time research powered by
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.
2140 experiments · showing 1261–1320
Test what an answer means, not just which words it contains, with Jev assertions for Vitest and Jest.
SOURCE SCREENSHOTEXPAND ↗
Pick a question or type your own, and Jev answers it twice: once alone, once with six live Exa results in its state.
ACTUAL RECORDING12 SEC ↗
choice
Six Exa results, each `Title — domain (date): highlights`, sit in `state.search_results` beside the typed question · yes / no
Fifty HR tool logos sit in a pile at the bottom of the page. Type an HR problem and the ones that fit rise into rows; Jev scores all fifty at once.
ACTUAL RECORDING12 SEC ↗
from the source
POST /api/jev/search {"q":"payroll for a small startup with eight employees"}{"picks":[{"id":"gusto","noul":0.91},{"id":"payfit","noul":0.71}],"ms":191}{"scale":{"choice":"smb","confidence":1}}
Four models play parallel Pong lanes; Jev chooses a paddle move whenever the ball needs another decision.
ACTUAL RECORDING12 SEC ↗
Watch Jev and a fly-connectome player race across separate Minesweeper boards, one move at a time.
ACTUAL RECORDING12 SEC ↗
Steer a Chicago traffic simulation with Jev choosing citywide priorities and bounded corridor policies.
SOURCE SCREENSHOTEXPAND ↗
Post in a simulated town and see how thousands of Jev-driven residents react to your message.
SOURCE SCREENSHOTEXPAND ↗
Type a color request and compare Jev’s ranked palette commands with an ordinary fuzzy match.
SOURCE SCREENSHOTEXPAND ↗
Make a tier list from your own items and criterion, with Jev scoring each item before code places it into S–F tiers.
SOURCE SCREENSHOTEXPAND ↗
Ask typed questions of a value with BAML’s .feels(), .how(), .matches() and .fill() methods backed by Jev.
SOURCE SCREENSHOTEXPAND ↗
Play a mahjong hand while Jev recommends a discard and rates its efficiency, safety and wait.
SOURCE SCREENSHOTEXPAND ↗
Compare Jev’s typed classifications with classical ML pipelines across the maker’s eight datasets.
SOURCE SCREENSHOTEXPAND ↗
Hide or highlight LinkedIn posts after Jev scores their writing against several slop signals and counter-signals.
SOURCE SCREENSHOTEXPAND ↗
Solve a bilingual lateral-thinking mystery by asking questions while Jev plays the keeper of the hidden answer.
SOURCE SCREENSHOTEXPAND ↗
Get a shell completion from your command history that changes as you type, rather than a static prefix match.
ACTUAL RECORDING12 SEC ↗
Watch Jev and a chat model judge a Fed chair’s remarks as hawkish or dovish, line by line.
ACTUAL RECORDING12 SEC ↗
Take penalty kicks against a Jev goalkeeper in a 3D football shootout.
SOURCE SCREENSHOTEXPAND ↗
Replay six Jev chess experiments, from unaided legal moves to tactical filters and Stockfish-assisted variants.
ACTUAL RECORDING12 SEC ↗
Turn a Telegram channel’s posts into a year-in-review card showing their mix of news, jokes, ads and more.
SOURCE SCREENSHOTEXPAND ↗
Write a browser test in Gherkin and let Jev map its steps to the controls on the page.
ACTUAL RECORDING12 SEC ↗
Tour a graphical Jev dojo with Tetris stacking, task routing, swarm reactions and other decision scenes.
SOURCE SCREENSHOTEXPAND ↗
Pitch a startup idea to a retro infomercial and compare how stupid Jev thinks it is with how fundable it looks.
SOURCE SCREENSHOTEXPAND ↗
Inspect listening ports on a Mac and ask Jev which processes to keep, clean up, or leave for your decision.
SOURCE SCREENSHOTEXPAND ↗
Let Jev steer a Solari browser or desktop: it picks actions from visible controls, and the agent checks each move before continuing.
ACTUAL RECORDING12 SEC ↗
Crawl a site to uncover orphan pages and get Jev-ranked internal links with suggested sentences and anchor text.
SOURCE SCREENSHOTEXPAND ↗
Race Jev through ten snap-judgment rounds, deciding whether reviews, messages, and other prompts are genuine, sarcastic, urgent, or invented.
SOURCE SCREENSHOTEXPAND ↗
Can a cheap classifier label every agent failure in a training run? Applied Compute clusters a sample of traces, then has Jev do the labeling.
SOURCE SCREENSHOTEXPAND ↗
from the source
Each model outputs a floating-point score from 0 to 1 for each of 14 failure modes.A score at or above the decision threshold marks that failure mode as present.
Give a chat app a few named subagents and let Jev decide who takes each turn and whether they work at once or in turn.
SOURCE SCREENSHOTEXPAND ↗
choice
Who must handle this turn? · main or an agent name · then parallel or sequence
Three Jev tools for the DeepSeek Harness on WSL: ask typed questions, check whether evidence supports a claim, or rank candidates.
SOURCE SCREENSHOTEXPAND ↗
from the source
return await client.systemOne({ state: args.state, questions })
3,608 one-subject-swapped statements put to Jev as forced true/false picks, in three wordings and eleven languages, with every raw answer published.
SOURCE SCREENSHOTEXPAND ↗
from the source
self.url, self.model = JEV_URL, "jev-latest"
A zsh plugin whose ghost-text command suggestions are ranked by Jev on demand, with shell history redacted before anything leaves the machine.
SOURCE SCREENSHOTEXPAND ↗
from the source
response = self.client.system_one(
A zero-dependency Node playground whose four tabs send typed questions to Jev through OpenRouter and draw the probabilities.
SOURCE SCREENSHOTEXPAND ↗
from the source
upstream = await fetch(ENDPOINT, {
Runs the CLASH cross-modal contradiction test against Jev with each image replaced by its COCO caption, then reports accuracy and modality bias.
SOURCE SCREENSHOTEXPAND ↗
from the source
r = self.session.post(f"{BASE_URL}/v1/systemone", json=body, timeout=self.timeout)
Every move in 2048 is one Jev Choice over the four directions, with no heuristic fallback and the probabilities, latency and cost shown live.
ACTUAL RECORDING12 SEC ↗
from the source
res = await fetch(url, { method: "POST", headers, body: JSON.stringify(body) })
Auto-marks 2,054 MR-GSM8K maths scripts against a teacher rubric with one Jev call each, then checks the marks against human annotators.
SOURCE SCREENSHOTEXPAND ↗
from the source
const result = await client.systemOne(
A CPU-first GUI agent where RapidOCR reads the screen, Jev picks the next action, pyautogui performs it and dHash checks the result.
SOURCE SCREENSHOTEXPAND ↗
from the source
r = _retrying_post(self.base_url, json=payload, api_key=self.api_key)
A local visual playground for System One APIs: build Yes/No, Choice and Score decisions, run them against several endpoints, and diff the results.
SOURCE SCREENSHOTEXPAND ↗
from the source
url=provider["endpoint"]
A companion that answers only with a face: Jev picks one of sixteen moods and blobatar morphs its expression, with no text reply.
SOURCE SCREENSHOTEXPAND ↗
from the source
const res = await fetch("https://api.typesafe.ai/v1/systemone", {
A Home Assistant integration that turns AI Task fields and YAML sensors into System One questions and decodes the answers back into Home Assistant values.
SOURCE SCREENSHOTEXPAND ↗
from the source
async with self._session.post(
Browser and computer use for coding agents where each step is a batch of small typed questions to Jev and code composes the next action.
ACTUAL RECORDING12 SEC ↗
from the source
response = await this.fetch(`${this.baseUrl}/v1/systemone`, {
A Chrome extension that covers YouTube comments until Jev decides whether, read against the video, they disclose a plot event.
SOURCE SCREENSHOTEXPAND ↗
from the source
upstream = await fetch(env.TYPESAFE_URL, {
Greek backgammon against Jev, which reads the position with four questions and picks the move as one Choice while code enumerates and prices every legal play.
SOURCE SCREENSHOTEXPAND ↗
from the source
upstream = await doFetch(UPSTREAM, init)
An autonomous Tibia bot: an OTClient Lua module streams health, mana, position and visible monsters every 250 ms and Jev decides survival, danger and targeting.
SOURCE SCREENSHOTEXPAND ↗
from the source
const response = await this.client.systemOne({
A Chrome extension that filters ads and uninteresting cards out of Xiaohongshu and Bilibili feeds, with Jev classifying each card and judging your interest in it.
SOURCE SCREENSHOTEXPAND ↗
from the source
const resp = await fetch(CONFIG.apiBase, {
A garage of small inspectable demos where Jev supplies the judgment — fraud screening, cyber triage, medical routing, emergency dispatch — and plain Python policy decides what to do.
SOURCE SCREENSHOTEXPAND ↗
from the source
return client.system_one(state=STATE, questions=QUESTIONS)
The open data behind an independent benchmark that puts Jev and six LLMs through the same typed questions on PubMedQA, Banking77 and HelpSteer2.
SOURCE SCREENSHOTEXPAND ↗
from the source
run header: {"type":"header",...,"model":"typesafe-ai/jev",...,"probability_source":
A Claude Code and Codex plugin that checks each tool call and final answer against the project's own rules with a System One model.
SOURCE SCREENSHOTEXPAND ↗
from the source
const res = await fetch(`${baseUrl.replace(/\/$/, "")}/v1/systemone`, {
The 2026 Korean CSAT Korean-language paper turned into a JSON state and typed choice questions, with Jev's answers, probabilities and error analysis kept alongside.
SOURCE SCREENSHOTEXPAND ↗
from the source
{\"1\": { \"type\": \"choice\", \"instructions\": { \"question\": \
Turns an image into the text state Jev can judge — local OCR in reading order plus a character-grid layout map — and returns exactly what the model will see.
SOURCE SCREENSHOTEXPAND ↗
from the source
return client.system_one(state=state, questions=questions, model=model)
A Neovim intent router that sends your prompt to Jev as one Choice and dispatches to a read, edit, shell or chat handler.
SOURCE SCREENSHOTEXPAND ↗
from the source
M.defaults = {endpoint = \"https://openrouter.ai/api/alpha/decisions\",model = \"typesafe/jev-1.13\",
Maps threat-report prose onto MITRE ATT&CK techniques and emits a Navigator layer you can load straight into the ATT&CK interface.
SOURCE SCREENSHOTEXPAND ↗
from the source
self._reflex = reflex or Reflex(model=model, lint=\"off\", guard=\"flag\", **reflex_kwargs
A local RAG workbench over Chinese League of Legends patch notes: every number comes from parsing the announcement, and Jev only reranks candidates and checks its own answer.
SOURCE SCREENSHOTEXPAND ↗
from the source
ENDPOINT = os.environ.get(\"TYPESAFE_ENDPOINT\", \
An independent benchmark of Jev 1.13 against a cheap and a frontier LLM on SMS spam and Banking77, measuring accuracy, ECE, Brier, latency and cost.
SOURCE SCREENSHOTEXPAND ↗
from the source
resp = requests.post(\"https://openrouter.ai/api/alpha/decisions\",
A zero-dependency local playground for Jev with a structured question editor and three presets: browser operation, intent routing and agent context trimming.
SOURCE SCREENSHOTEXPAND ↗
from the source
const API_URL_PRESETS = {official: 'https://api.typesafe.ai/v1/systemone',openrouter: 'https://openrouter.ai/api/v1/systemone',
A creator-inbox desk that asks Jev five questions per comment and lets code choose the action; a confidence slider re-routes the whole queue without another call.
SOURCE SCREENSHOTEXPAND ↗
from the source
const result = await client.systemOne({state: { ... },questions: QUESTIONS,
A drop-in proxy that answers your app with Jev byte for byte while asking the same question of local doubles, then reports where they would have decided differently.
SOURCE SCREENSHOTEXPAND ↗
from the source
\", \"headers\": { \"authorization\": \"Bearer ${TYPESAFE_API_KEY}\" }, \"model\": \"jev
A Next.js workbench for Jev's three primitives through OpenRouter: a 95-category Choice, Noul evaluation with transfer curves, and Score rubric rating.
SOURCE SCREENSHOTEXPAND ↗
from the source
const endpoint =
A web agent with no LLM in it: each step asks Jev five typed questions in one call, and code gates the answers before anything is executed.
SOURCE SCREENSHOTEXPAND ↗
from the source
questions: dict[str, Question]
A Chrome side panel that auto-trades an NSE stock through Zerodha Kite: Jev answers a direction question each tick and a risk gate sizes and places the order.
SOURCE SCREENSHOTEXPAND ↗
from the source
const res = await fetch(`${JEV_BASE_URL}/v1/systemone`, {method: \"POST\",signal: ctrl.signal,headers: { \"Content-Type\": \"application/json\", Authorization: `Bearer ${apiKey}` },body: JSON.stringify({ state, model: JEV_MODEL, questions }),
Has Jev play chess against itself one decision per move, with a browser board showing the position and Jev's last answer.
SOURCE SCREENSHOTEXPAND ↗
from the source
endpoint = f\
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 experimentSign in first so we can reply to you. Submissions are reviewed, not published automatically.