Research & data

A worked guide to Jev's Score primitive: writing a request, defining rubric levels, reading recorded probabilities and the weighted-score math.

Sends a detected bot's last 48 request paths to Jev, which picks what it's after (prices, articles…) and how it crawls (pagination, IDs…), or unknown.

Dan Semenescu

@DansiDanutz

↗

Paste an article or post link. Fake / Real extracts its claims, finds outside evidence, and has Jev judge whether the evidence supports or contradicts each one.

Combines rubric science and LLM-as-a-judge research to grade outputs with AI judges: LLMs, decision models like Jev, or both.

Adnan

@_itzadnan_

Another Jev use case - Dreaming! I tested two different Dreaming pipelines: A purely LLM-based pipeline using Gemini. A hybrid pipeline using Jev for classifying and maintaining knowledge graph. Check it Out - https://t.co/24x0rqZGLy Repo -

Johnny Makhoul

@johnnymakhl

Stop letting LLMs or Jev write raw SQL against production data or even evaluate rows. I built a CLI that routes questions through a semantic layer (Cube/dbt) with confidence checks so it only runs safe, verified queries. Quick demo running on: https://t.co/ACoN1ld7MM

priorbench

github.com

↗

A preregistered evaluation publishes request code, raw responses, and outcomes across 21 experiments on Jev decision behavior.

simonmesmith

github.com

↗

A frozen Jev Choice setup classifies 3,080 banking support messages using category definitions and retrieved labeled training examples.

PistachioAIHQ

github.com

↗

Jev combines include/exclude choices with atomic eligibility checks to screen research abstracts against a published systematic-review dataset.

A review-first prototype uses Jev to classify chip parameters and judge whether quoted datasheet evidence supports an engineering claim.

libingzheren

github.com

↗

Jev connects memories, routes retrieval across graph views, scores candidate evidence, and decides when the answer model has enough context.

kitfunso

github.com

↗

An opt-in Jev reranker scores candidate memories from a local SQLite store before selected evidence reaches a coding agent.

Gaurav-Gosain

github.com

↗

A reproducible security evaluation uses Jev probabilities to detect prompt injection and vulnerable code on public labeled datasets.

Kamimoto(かみもと)

@sep_is_heim

流行のJevをMiniMax H3に組み込んで、動画生成を高速化してみた！Attention処理のスパース化にJevを使用。 ・層ごとにJevが重要度を判定（4step 49層が対象） ・Jevがスパース率1%, 3%, 5%, 10%を選択 RTX4070で6分7秒→3分34秒で41.7%短縮！動画生成中にJevクラウドに問合せしているのに速い！

Lucian in sf

@xucian_

I made Jev talk and wrote a paper on it. To publish it on arXiv, I need an endorser for csCL fom someone who has 3+ arXiv papers in any cs category, submitted between 3 months and 5 years ago. It's one click after I DM you the code and the paper. Or maybe someone you know. A

Jσsh

@JoshKuechly

I asked Jev to predict its own future. My last post (below) got 300k+ impressions, 400+ comments, 101 reposts. I used Jev to categorize every comment and quote to capture public X opinion on Jev’s future. Here’s results: ———— *Note: I created five prediction buckets,

Christopher Lee

@chris_not_busy

I built an ultra-fast medical simulation with Jev and Grok 4.7 Grok 4.7 and I co-doctored a cancer patient across 20+ possible futures in just minutes. Grok proposed the moves. Jev ran the high-speed state decisions < 0.3 secs Patientic visualized the branched outcomes. 🧵👇

Ankita Tripathi

@ankitatr_

Built a small experiment to understand my chess beyond “blunders” and “accuracy.” https://t.co/vJlJFW03dj games → Stockfish for objective move analysis → Jev for recurring semantic patterns → code for trends and loss/win comparisons. Now the dashboard can show what keeps

Andy

@andywang

The bookkeeping services industry is dead. This weekend, I built a better solution using Jev from @typesafeai. I fed it 34 months of work a firm charged $20,000+ for. Jev did a better job in 20 seconds, for just $0.32 🤯

Dhruv

@dhruv_ko

1/ Jev has a chess rating now. 1023 Elo. Jev is TypeSafe's "System One" decision model. No search, no planning. You hand it a situation and options, it picks one with a confidence score. Built for routing and classification. Not games. So I built it a chess harness 🧵

Samridh Srivastava

@CallmeSamridh

Built a music intelligence platform with Jev. 3,274 songs. 114 genres. Measured “Soul” vs “Subversion.” And the entire audit cost me ~$0.0001. That’s the part that blew my mind. Subjective intelligence at massive scale, for basically nothing. #Jev #AI #MusicTech

Mikhaeel

@mmmikhaeel

I built a tool powered by Jev that extracts any products features and business logic into a markdown repo Enter a product URL and you can copy features, ICP, messaging, pricing strategy, etc straight to a md file

Shrithan

@DevaiahShrithan

I gave JEV at every AI paper since 1993

An open-source app curates Hacker News stories from a user’s stated interests using Jev.

An open-source app uses Jev to score more than 3,000 emojis against any typed phrase.

jevals turns agent checks into typed questions so the same evaluation definitions can run on saved traces and as runtime guardrails with Jev-style decision models.

A custom Jev harness uses context-sensitive hints and chained decisions to pronounce Chinese characters.

Jev helps find the exact moment answering a health question across more than 3,000 hours of video.

A Chrome extension uses Jev to choose the best existing prompt, template or workflow for the user’s goal even when their search terms do not match saved titles, text or tags.

Rebuilt a survey-research workflow with Jev for 200 samples and 12 questions per survey; author reports it ran 2x faster and cost 85% less than Gemini 3.5 Flash-Lite.

TypeSafe launched Jev mid-September 2026. It makes decisions, cannot write a single sentence, and is 20 to 200x faster and 40 to 400x cheaper than frontier models according to…

TypeSafe AI (Jev) decisions for Pi: batched Choice/Score/Noul evaluation tool, terminal playground, and a typed API other extensions build on. ... Install `pi-typesafe` from npm…

LLMs Are Built to Write. Jev Is Built to Decide. | by Dr. Fadi Shaar | AI Mindset | Sep, 2026 | Medium ... # LLMs Are Built to Write. Jev Is Built to Decide. ... Inside Jev,…

The simplest way to write decision workflows in Python, backed by TypeSafe's Jev model ... classifier, decision, jev, routing, transpiler, triage, typesafe ... ## Description…

Jev: a decision-model experiment (TH) ... # Jev: a decision-model experiment (TH) ... 1. ### Jev ฉลาดพอไหม? - บททดสอบความสามารถของโมเดล "ตัดสินใจ" (ตอนที่ 1) ... Jev คืออะไร?…

That is almost exactly the territory TypeSafe claims Jev is built for. ... The narrowness of the test is important here because it relates exactly to what TypeSafe claims Jev is…

That question is already being answered for Jev, the first public System One Model from TypeSafe AI. ... ## What is Jev by Typesafe AI? ... TypeSafe AI describes Jev as AI built…

JEV is TypeSafe's System-One model (`jev-latest`), served at `api.typesafe.ai/v1/systemone`. It's text-only. No vision, no market opinions. You POST it a "state" string plus a…

After two years in stealth, Diogo Almeida, one of the researchers behind ChatGPT, launched Jev at TypeSafe AI. The company advertises roughly 200 times faster and 400 times…

That gap is exactly what Jev AI is built for. Jev is a System One model from TypeSafe AI that returns typed decisions with probabilities instead of generated text. You send it…