Case wall
Jev AI examples, page 19
615 real Jev AI builds and public demos, updated daily. Original text, metrics, and a link back to each source.
Last updated
Public Jev AI examples and demos, not official TypeSafe docs. Clients and SDKs are in the Jev GitHub projects. Paths for jev-latest, Vercel, and OpenRouter are in the September 2026 channel map.
Agent-skill repositories are listed on Jev skills on GitHub.
OpenRouter model ids recorded for Jev AI are on Jev AI on OpenRouter. Rubric and eval posts are on Jev as a judge. How the company and the model fit together is on TypeSafe AI. The pricing calculator uses the published input rate.
No posts match these filters.
@akira_papa_IT built a chat app by making Jev judge the next character. They say it is very fast, input is $0.042 per 1M tokens, and output is free without a cap.
Jevで次の文章予測判断させるロジック作って、 判断特化AIのJevで、AIチャットできるアプリ作ったんだけど、まず爆速だし、入力は$0.042 / 1Mトークンだけど、出力が無制限で無料でした。 出力無料のAIチャットってなんだこれw 「え?これ作れちゃっていいん?」ってなってる。。。 #Jev https://t.…
@CompleteSkeptic, ChatGPT co-inventor Diogo Almeida, announces Jev after two years building a training method called RLCD. The launch claims 20–200x faster, 40–400x cheaper with free output tokens, and frontier composable intelligence optimized for decisions.
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to t…
Quoted @thread
@coolish compared playing Slay the Spire 2: GPT-6 Astra was capable but slow, while Jev acted in 0.7 seconds, fast enough that they say they could not follow the screen.
昨天出的 https://t.co/6eiMFaJTrj Jev 模型 适合高频决策,那岂不又是适合打游戏? 之前用 GPT-6 Astra 代打杀戮尖塔2 能力强但速度慢。 我刚实测用 Jev 打,行动思考只需 0.7秒, 画面我都没看清它就操作完了。 这超人类游戏速度, 着实又让我震惊瘫坐了😅 语言很难描述这感觉,…
Quoted @CompleteSkeptic
@nikunj says Jev scored 3,000 kid snacks on nosugarforkids.com against multiple criteria in 28 seconds for $0.11.
Jev scoring 3000 kid snacks with multiple criteria on https://t.co/KSiodPdZHT in 28 seconds for $0.11 🤯 https://t.co/c9O6uTyiUK
Quoted @nikunj
@jpschroeder says they rebuilt Tesla Full Self-Driving with Jev in under an hour. The post states that claim and links a clip; it does not describe the sensors, stack, or what the demo actually drove.
I rebuilt Tesla Full Self Driving with Jev in less than an hour. This model is a total unlock. https://t.co/mUrvaFrlQ7
Quoted @thread
@tsuyoshi_osiire prototyped a sensor that watches a conversation in real time with Jev, updating explore versus converge, resolved versus open, and disagreed versus agreed after each utterance, for uses such as live user-interview analysis.
Jevで、会話をリアルタイムに観測するセンサーを試作しました。 発言のたびに ・探索 ↔ 収束 ・解消 ↔ 未決 ・未合意 ↔ 合意 を更新し、会議の状況を可視化します ユーザーインタビューのリアルタイム解析などにも使えそう。 https://t.co/cO21f4AK99
Jev chooses an action from the accessibility tree, then Stagehand executes it.
we built blazing fast computer/browser use with Jev + @Stagehanddev. this task cost $0.001 and executed at near instant speed (in a remote browser btw) the loop…
@reczko_konrad combines TypeGPU, ruNNtime, and TypeSafe Jev: local inference and rendering share the GPU with zero copy, three neural nets plus rendering run in real time, and Jev adds the semantic step from camera and mic (Moonshine, YOLO26, DepthART) to lights, shadows, and bloom.
TypeGPU + ruNNtime + Jev @typesafeai is a very fun combo :D ruNNtime gives me efficient local inference, TypeGPU lets inference and rendering share GPU resource…
@rinte0321 demos slides that appear automatically from what is being said, so the speaker does not touch the PC or a pointer. They say the speed and price are what make that possible.
jevデモ② 発話内容に応じたスライドの自動表示。 喋った内容にマッチしたスライドを自動で表示してくれます。PCやポインターを触らずにページを表示できるので、これで誰でもジョブズになれるかも?? 速い・安いだけでできることがこれだけ広がるのかと感動している。 https://t.co/3UO6DQcbJ2
@MatijaSosic posts a 45-second TL;DR of Jev. They say the core idea is simple, but another video made it hard to understand.
Here's a 45-second TL;DR on Jev. I find the core idea beautifully simple, but the video made it really hard to understand. Hope you find it helpful. https://t.c…
Quoted @CompleteSkeptic
@hi_im_isaac_ says Jev cannot generate text, then gave it a few hundred common English words plus punctuation to choose from so it appears to say something.
"jev can't generate text" But jev has a lot to say! I gave him a couple hundred common english words to choose from + punctuation https://t.co/WmDkTyiG9u
@sydneyrunkle says Jev from TypeSafe returns structured output instead of generated text, which suits classification such as model routing, tool selection, and guardrails, and that it is fast and cheap compared with LLMs on those tasks.
instead of generating text, jev from @typesafeai generates structured output this makes it great for classification tasks like model routing, tool selection/sea…
@ytiskw asked 150 fictional personas 12 questions each about whether they would adopt a product. They report about 1.8 yen and about 5 seconds, including Japan–US latency.
話題のJevで架空のペルソナ150人に製品の導入意向を聞く仕組みを作ってみた。150人ごとに12個の質問をAPIで投げる。 実際にかかった費用は、1.8円でかかった時間は約5秒。十分爆速だけど日本↔︎アメリカのレイテンシが含まれているので本来はもっと早いはず https://t.co/6lbFXY9jjH
@nutlope classified 100 emails (50 legit, 50 fraudulent) with Jev in 1.42 seconds, then sent the 31 predictions under 95% confidence to Kimi K3. The combined pipeline was 96/100 correct in 16 seconds for about $0.07, of which about $0.003 was Jev.
Jev + Kimi K3 for fraud detection! TLDR: Jev classified 100 emails in 1.42 seconds, then I routed the uncertain cases to Kimi K3. The full pipeline got 96/100 c…
@aimlapi had Jev V13 play 5+0 blitz chess, one API call per move, against Fable 5.1 and GPT-6 Astra. They report Fable was up a queen and +16 material by move 29 but lost on time because it spent 6–15 seconds per move while Jev answered in about 2.6 seconds; Astra checkmated Jev in 18 moves (Qe1#) with 2:27 left.
Jev VS Fable 5.1 VS GPT-6 Astra: chess typesafe's new Jev V13 isn't an LLM. it doesn't chat, doesn't explain, doesn't write code — it only makes decisions. so w…
Quoted @CompleteSkeptic
@savboj posts that Jev plus computer use is 100x faster than any LLM, and that if you blink you might miss it. The text does not describe the task.
If you blink you might miss it Jev + Computer use = 100x faster than any LLM @typesafeai https://t.co/JG09Wx7UCo
Quoted @CompleteSkeptic
@nutlope classified 1,018 AI papers: DeepSeek V4 Flash summarized them for $3.99, then Jev classified title plus summary into 24 topics for $0.08 total, with 256ms median end-to-end latency per paper. The site 1kpapers.com is live, and they say evals are still running before the Jev labels replace the current ones.
I used Jev to classify 1,018 AI research papers. The result: $0.08 total cost and 256ms median end-to-end latency per paper. The pipeline was: 1. Summarize each…
@JonPTaylor shows Jamcat, a voice UI where Pipecat hands off between a Jev command agent that drives the UI and PhoneLLM for generations and discussion. They are in London calling us-west and still call it snappy.
Wow, @typesafeai Jev is a game changer for voice-driven UI. Responsiveness really matters for an experience to 'feel' right, and this certainly does the job! La…
@AlanDaitch says Claude wired Jev to Playwright to search used listings. It reads about 26 listings a minute, decides in 406ms, discards poor fits, bids on matches, and messages sellers when a field is missing. They say the search cost $0.00085, or about 26,000 listings per dollar.
Claude me integró Jev con Playwright para buscar usados. Lee unos 26 artículos por minuto y decide qué hacer con cada uno en 406 milisegundos Descartó los que n…
@hot_town posts an explainer covering how Jev differs from an LLM, how it works, real-life examples, and when to use it.
Jev is here. - how is different from an LLM? - how does it work under-the-hood? - what are some real life examples? - when should I use it? I answer all that be…
@wuyang_zhou posts what happens if you ask Jev to play Minecraft together with GPT-6 Astra.
btw this is what you get if you ask Jev to play Minecraft with GPT-6 Astra. I still can’t believe it https://t.co/ERDnsVWPLZ
Quoted @CompleteSkeptic
@SUOHA_AI posted a 50-second video explaining Jev and its architecture: a System 1 model from Diogo Almeida after two years, one parallel pass instead of token-by-token generation, RL for calibrated decisions, and typed outputs. They say the video also shows an 8.5-second autoregressive delay next to a real-time minigame, referencing a @MatijaSosic demo.
很多人没看懂爆火的JEV是什么? 所以我做了一个 50 秒的演示视频,带大家把海外刷屏的 JEV 以及它背后的底层架构看透👇(见下方视频) JEV 是前 OpenAI、参与 ChatGPT / RLHF 的 @CompleteSkeptic(Diogo Almeida)创办 @typesafeai 后的第一款模型 他在…
Quoted @SUOHA_AI
@marcelpociot asks whether TypeSafe Jev will turn $10,000 into profit and shows paper trades whose decisions run through Alpaca.
Will Jev @typesafeai turn $10.000 into profit? Let's find out! Running decisions for paper trades via https://t.co/28WVX50sv3 https://t.co/2XglDmGTNM
@Neel490 says the implications of Jev for self-driving could be huge and links a clip. The text does not describe a system.
I think the implications of jev on self-driving could be huge https://t.co/advoRDSOz7
Quoted @CompleteSkeptic