Skip to content

Latest commit

 

History

6 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

JEV HUB · Jev 聚合站

X(Twitter)上关于 TypeSafe AI「系统一模型」Jev 的长文与演示视频聚合。 逐条保留原帖链接与作者署名,本站不转载、不二次托管任何内容。

→ 打开聚合网页浏览全部内容 | 结构化数据 data/posts.json

由 谁是专家 出品(小红书 / 微信公众号 / X 同名)。

数据概览

指标 数值
动态总数 982
演示视频 382
长文 147
讨论 453
参与作者 818
语言分布 英文 673 · 中文 192 · 日文 106 · 韩文 6 · 阿拉伯文 5
数据跨度 2026-09-15 18:17 ~ 2026-09-21 03:59(UTC)
最近更新 2026-09-21 05:22(UTC)

🐣 小白入门(先读这篇)

读不懂也要读懂系列01|GPT它爸,造了个不会聊天的 AI

写给完全不懂的人看的 Jev 科普:从「一个不会说话的 AI 有什么用」讲起,把 Jev 为什么快、为什么便宜、能干什么活一次讲清。小白先读这篇,再看下面的演示和长文就不迷路了。

— @dboy_yi2025 · 2026-09-18

官方入口

⭐ 编辑精选(按互动量)

  • 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 ne… — @CompleteSkeptic · 2026-09-15 · 模型原理 · 成本速度
  • found the perfect use case for @typesafeai Jev: instant compaction in 2026, why is compaction still a summarization prompt? Jev can make it instant by… — @tamarajtran · 2026-09-17 · 智能体
  • The gains aren’t free: Jev can't generate text Comparing Jev vs LLMs side-by-side makes the trade-off clear Fun fact: replacing sequential computation… — @CompleteSkeptic · 2026-09-15 · 编程开发 · 成本速度 · 评测对比
  • 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. — @MatijaSosic · 2026-09-16 · 编程开发
  • Jev is now available to everyone. No waitlist. Start using it here: https:// console.typesafe.ai — @typesafeai · 2026-09-20
  • Jev was adopted faster than any other model in AI Gateway history. In the first day, @typesafeai reached ~13% of teams, 2x the GPT-5.6 family and 6x F… — @vercel · 2026-09-18 · 成本速度 · 评测对比
  • 【Breaking News】ChatGPT Co-Inventor Unveils New AI Model "Jev" After 2 Years of Stealth Development ・20-200x faster than LLMs, 40-400x cheaper ・Output … — @ctgptlb · 2026-09-16 · 成本速度 · 评测对比
  • jev is INSANE. in 40 seconds it broke down 724 live ads from 37 brands. every hook. every format. offer. cta. awareness stage. landing page mismatch. … — @TheMattBerman · 2026-09-17 · 成本速度
  • we are officially out of stealth! join the frontier and get access to Jev on our website (link on profile) — @typesafeai · 2026-09-15
  • I created an EC real-time customer service demo using Jev and gpt-live-1. Even in the middle of a conversation, it immediately suggests recommended pr… — @rinte0321 · 2026-09-17
  • New experiment: json-render + jev The future Generative UI is instant Your components, your actions, your design system Rendered in milliseconds — @ctatedev · 2026-09-18 · 编程开发 · 成本速度
  • Jev by @typesafeai is now on OpenRouter, in beta. Jev is a System One model. Instead of generating text, it takes your app's state plus a typed questi… — @OpenRouter · 2026-09-18 · 编程开发 · 模型原理

🎬 演示视频精选(Top 60)

  • 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 ne… — @CompleteSkeptic · 2026-09-15 · 模型原理 · 成本速度
  • found the perfect use case for @typesafeai Jev: instant compaction in 2026, why is compaction still a summarization prompt? Jev can make it instant by… — @tamarajtran · 2026-09-17 · 智能体
  • 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. — @MatijaSosic · 2026-09-16 · 编程开发
  • jev is INSANE. in 40 seconds it broke down 724 live ads from 37 brands. every hook. every format. offer. cta. awareness stage. landing page mismatch. … — @TheMattBerman · 2026-09-17 · 成本速度
  • we are officially out of stealth! join the frontier and get access to Jev on our website (link on profile) — @typesafeai · 2026-09-15
  • I created an EC real-time customer service demo using Jev and gpt-live-1. Even in the middle of a conversation, it immediately suggests recommended pr… — @rinte0321 · 2026-09-17
  • New experiment: json-render + jev The future Generative UI is instant Your components, your actions, your design system Rendered in milliseconds — @ctatedev · 2026-09-18 · 编程开发 · 成本速度
  • Jev by @typesafeai is now on OpenRouter, in beta. Jev is a System One model. Instead of generating text, it takes your app's state plus a typed questi… — @OpenRouter · 2026-09-18 · 编程开发 · 模型原理
  • 完全睡不着了....这个视频让你完全看懂JEV的恐怖能力 我实际演示了一下,用 JEV 和 DeepSeek 做同一个任务:从当天的海量实时新闻流里,为 15 个品牌快速匹配有没有适合借势的公关热点,并打上结构化意图分类标签 结果是:28 秒内,Jev 狂刷完了整整 428 — @SUOHA_AI · 2026-09-18
  • 💥炸了!开源本地版 Laya 响应速度吊打 Jev ,毫秒级决策快到飞起! 左边 Laya 是本地 421M 开源决策模型,右边 Jev 1.13.0 是云端 API。 同一盘贪吃蛇、同一套 typed decision,30 秒自由跑下来: 🔹Laya:分数 46、长度 52,决策 86.5 次/… — @NFT_Chen · 2026-09-20 · 编程开发 · 成本速度
  • 昨天出的 https:// typesafe.ai 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… — @coolish · 2026-09-17
  • Browser Use 接上 JEV 这速度太他妈吓人了吧?? 过去大家看 GPT、Claude 操控浏览器,体验其实并不好,因为真的太慢了,中间还经常卡死... 因为让大模型写一堆长篇大论去操作网页,本质上就是杀鸡用牛刀 在网页上点按钮,本质上就是一个离散的多选题,根本不需要 AI 会写诗 JEV… — @SUOHA_AI · 2026-09-17
  • 卧槽,居然真有比Codex内置computer use更快的操作电脑的方式 我尝试用Codex+Jev打造了一个加强版computer use,我称之为「Jev Use」。比内置的更快更丝滑,token消耗却差不多 下面是我用「添加Mac日历事件」做了一个对比视频,同样的内容,明显可以看到Jev版本… — @Saccc_c · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • JEV captcha arbitrage sounds insane... If a Chinese marketplace pays $0.01 per CAPTCHA and JEV can solve 100 for $0.0068, that’s a 99.32% gross margin… — @kenonews · 2026-09-20 · 成本速度 · 金融交易
  • @OpenRouter 🤝 @typesafeai — @typesafeai · 2026-09-18
  • 直观感受下 seo 内链重构:jev pk opus5 Jev 在 45.1 秒内完成了全站重构,成功放置 584 个高相关性内链,并拒绝了 139 个强行凑数的无效页面,总耗费 $0.21。 作为对比,opus5 相同时间只处理了 21 个页面。 — @ai_xiaomu · 2026-09-19
  • 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… — @SUOHA_AI · 2026-09-18
  • 这也太夸张了吧!学到了!! 这拿来做竞品分析,简直有如神助!! 37 个品牌、724 支广告,40 秒拆完! 跑完一轮分析,成本0.09美元!! 做竞品分析的朋友,看到这个真的很难不心动。 Jev 批量分析竞品广告,系统拆解背后的思路,从开头怎么吸引人,到最后怎么促成行动,覆盖六个维度: — @yyyole · 2026-09-18
  • 我用 @typesafeai JEV 模型做了一个实时 3D 场景生成器。 这个速度实在太夸张了! 从预制的几十上百种 3D 模型里,根据输入的文案,进行并发几百次的判断。 同时对这些素材的着色、光照、位置以及状态进行处理,一秒完成符合要求的室内场景搭建。 — @op7418 · 2026-09-20
  • Jev is incredible. I'm scared people are using it wrong. — @theo · 2026-09-21
  • 💥我靠!用GPT-6 Astra把Jev运行结构可视化,直接把 LLM 和 JEV 的差别钉死了! 同一句:“网络一直连不上,请帮我处理。” LLM:一边在讲故事,一边在做判断。 🔹逐词元往下写 🔹前文决定下一词 🔹最后吐出一句完整话:「建议转给技术支持处理。」 🔹软件还得再解析这句话 — @NFT_Chen · 2026-09-18
  • 💥实力比拼:Jev vs DeepSeek V4.1 Flash,瞬间秒懂Jev的恐怖能力! 500条真实电商客服工单,同一套分单任务: 🔸JEV:83秒清完,花了 $0.01,左边已经全部归档。 🔹DeepSeek:这轮停单时只做完 173条,花了 $0.06,右边还剩一大半,贵了大约5倍。 差距… — @NFT_Chen · 2026-09-19 · 评测对比
  • 💥我靠!这是目前看到的给 Jev 模型找到的最狠的真实用例! Instant Compaction只删,不改写。 演示里 156k tokens 的 Claude Code 会话,Jev 逐条给 tool call 打分,瞬间压到 62k。 用户目标、精确报错、关键代码原文全在;过期 Read / … — @NFT_Chen · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • 终于看到了大语言模型之后的一个新的模型范式 Jev 是系统1模型,极速响应,价格超低,但不经过语言系统,不能说话,完全客观,也不需要语言层面的对齐。 潜力很大,也许是通向 AGI 的另一条路。 — @oran_ge · 2026-09-17
  • 最近 Jev 挺火,我拿它去 12306 查了趟动车票,看看这种浏览器 Agent 到底能不能干点日常的活。 用的是 Browser Use 开源的 Jev Ultrafast,一个主打速度的浏览器 Agent,可以自己操作网页,完成你交给它的任务。 我这次用 Pi + DeepSeek — @yanhua1010 · 2026-09-19 · 智能体 · 成本速度
  • Jev by @typesafeai is now on the Venice API, in beta. It answers, it doesn't write. Send it your app's state and typed questions. Get back a typed ans… — @AskVenice · 2026-09-18 · 编程开发 · 模型原理
  • Someone just open sourced jev-trader, a fully working 24/7 trading bot with Jev that answers buy or sell on every block with calibrated probabilities … — @RoundtableSpace · 2026-09-21 · 编程开发 · 模型原理 · 成本速度 · 工具资源 · 金融交易
  • 强烈建议大家去看一下 Jev 创始人 Diogo Almeida 的这个演讲,关于 Jev 的一切,他在这个视频中早就预言。 Jev 一开始就是盯着自动化去的,Diogo Almeida 认为现在的 LLM 陷入了一个史诗级大弯路,就是 RLHF。 RLHF — @threeaus · 2026-09-20
  • 太爽了!终于找到了JEV的最强应用场景之一--全自动化 Browser use 我用JEV全自动填完 TypeSafe AI 的申请问卷,我只给它丢了一个完全陌生的网址,让它全自动帮我完整填完:38 秒一次性刷完 16 道题,全程没停顿、零人工干预 具体的逻辑是: 我看到 @typesafeai 文… — @SUOHA_AI · 2026-09-20
  • Jev combined with Grok Bot makes the agent setup cheaper and faster than 95% of what people are running, with Grok Bot handling execution while Jev ma… — @RoundtableSpace · 2026-09-21 · 智能体 · 成本速度 · 评测对比
  • Laya-MLX is an open source classification system similar to Jev but 50x faster, running locally on device with a maximum of 1GB memory and making deci… — @RoundtableSpace · 2026-09-21 · 编程开发 · 游戏对战 · 成本速度 · 内容审核 · 工具资源
  • someone finally fixed the expensive part a new AI called Jev was released only a few days ago. it doesn’t generate paragraphs - it makes typed decisio… — @mika_systems · 2026-09-20 · 成本速度
  • This looks promising. It seems like it could significantly reduce usage costs. Claude Code model routing mod "jev-model-router" Automatically routes l… — @masahirochaen · 2026-09-20 · 编程开发 · 智能体 · 模型原理 · 成本速度 · 内容审核
  • A fully autonomous real-time trading bot built with Jev in one evening and morning, ingesting onchain and offchain data for rapid decisions, has lost … — @RoundtableSpace · 2026-09-21 · 编程开发 · 金融交易
  • What is a decision model? Jev by @typesafeai answers yes/no and multiple-choice questions, with a confidence score. Much of software development are a… — @OpenRouter · 2026-09-18 · 编程开发 · 成本速度 · 评测对比
  • Jev 太离谱了。 只用 40 秒,它就拆解了 37 个品牌正在投放的 724 条广告。 每条广告的钩子、形式、Offer、CTA、用户认知阶段,以及广告与落地页不匹配的问题,全都分析了出来。 Token 成本仅 0.09 美元。即将接入 StealAds + MCP。 — @FinanceYF5 · 2026-09-20 · 成本速度
  • 卧槽! 我把 Jev 玩到了 Grok Bot 里 跑 X 的时间线浏览, 看有没有贴子适合自己回复下, 这可是大大的省时间了 整个流程十分简单, 照着文章操作几分钟就能用起来 — @0xlangeai · 2026-09-20
  • This is what happens when a designer gets access to Jev. — @RoundtableSpace · 2026-09-21
  • I was super intrigued when I noticed that JEV returns not just probs for candidate choices, but also a CONFIDENCE SCORE of it's prediction... I was wo… — @neural_avb · 2026-09-20 · 编程开发
  • TypeSafe's judgment-specialized model "Jev" has released its fully open version "DJev." It also handles images reliably. ・Based on Google's DiffusionG… — @masahirochaen · 2026-09-20 · 编程开发
  • Jev + GrokBot is the best GTM system I've ever built It just made my GTM x200 CHEAPER and x400 FASTER than what 95% of teams are running... setup take… — @0xMovez · 2026-09-20 · 成本速度 · 评测对比
  • 🔥 卧槽了,jev 简直太适合 codex 用户了,强烈建议 codex 用户去尝试下 jev 因为 jev 非常适合需要快速进行复杂决策和评估的场景,所以和 codex 的 computer use 功能简直是绝配。 直接把 jev 用作判官,指导 codex 快速操作你的电脑,既准确又节省 to… — @alin_zone · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • People use Jev to pick a model before a task. I made it change GPT-6's reasoning effort inside Codex DURING the task. More thinking when stuck. Less f… — @miu21590 · 2026-09-21 · 编程开发 · 成本速度
  • 最近应该都被Jev 刷屏了。 但是该怎么玩呢? 一个视频教你在 WorkBuddy 中接入 Jev。 接入完该拿它来干什么呢? 它不会聊天、不会写代码、一个字都吐不出来。 只回答三种题:选哪个、打几分、是或否。 虽然大概率给的答案是正确的,但不代表不会出错。 所以最适合它的,不是「替代大模型」。 — @aehyok · 2026-09-19
  • Tested Jev vs GPT-5.6 for predicting my taste in books. I used my 1,000 Goodreads ratings as starting data and held out 100 to test. Then I asked both… — @venturetwins · 2026-09-19 · 编程开发 · 成本速度 · 评测对比
  • 用AI一定要试试Jev :跳过聊天,直接输出带置信度的结构化决策。特定任务快193倍、便宜445倍;开源插件能快速压缩上下文,长任务告别越跑越“失忆”。 Jev上手很简单: 1、去 https:// TypeSafe.ai 申请资格 2、让 Codex 安装 TypeSafe Skill 3、在后台… — @KKaWSB · 2026-09-18 · 编程开发
  • I had them create a demo of an AI assistant that reads the context of the conversation without an activation word and only responds when necessary, us… — @npaka123 · 2026-09-21 · 编程开发
  • Theo's opinion on Jev (Type Safe AI) — @maria_rcks · 2026-09-19
  • We built a plugin that gives Jev a browser in Cline, and have been blown away by the results. 1. Install it in our new desktop app: Customize > Market… — @cline · 2026-09-18 · 编程开发 · 工具资源 · 金融交易
  • Jev just solved doomscrolling. You pick a niche, Jev reads 3 days of posts and asks 8 questions each. It runs in ~2s for $0.007 😅 > removes bait and h… — @robj3d3 · 2026-09-18 · 游戏对战 · 工具资源
  • Jev dropped the price of SEO/GEO fixes by 90% Agents that audit and fix a client's SEO/GEO used to cost us ~$250 Here's where the savings come from: 1… — @irabukht · 2026-09-18 · 智能体 · 成本速度
  • 卧槽, 我让 Jev 新模型来分析下 Claude 封号问题,它居然全都判断正确了…… 这个决策太强了啊!! Jev 这不是还能做意图识别? Jev 这是从 OpenAI、A/ 中杀出一条血路…… — @xmglab · 2026-09-21
  • A Brief Introduction to the New JEV Model with Persian Subtitles This video examines the new Jev model, introduced by TypeSafe. Unlike common language… — @iSegar0 · 2026-09-20 · 编程开发
  • Looking for inspiration on what you are able to do with Jev (by @typefaceai ) this weekend? Built a website called Jevable ( https:// jevable.com) whi… — @nikunj · 2026-09-18 · 内容审核
  • Here's a 3-minute breakdown of "Jev," the hot topic buzzing on SNS right now. It's often hyped up with talk like "It surpasses LLMs!" but that's a mis… — @masahirochaen · 2026-09-18 · 成本速度 · 评测对比
  • TypeSafe 刚放出 Jev,立刻就有人搓了个独立站 https:// jev-ai.pro。玩法很眼熟:官方模型一出,第三方先把 playground 和付费入口搭好。 Jev 官网早期访问门槛在那儿,普通人想先玩一把、或者绕开排队调 API,就会撞上这类站。 https:// jev-ai.… — @0xlangeai · 2026-09-20 · 编程开发
  • Jev, that's way too insane, right... I pulled 300 cases from the 73,999 vehicle complaint reports publicly released by the US NHTSA and made a 3D demo… — @KanaWorks_AI · 2026-09-20 · 智能体 · 内容审核 · 工具资源
  • It seems that Jev is also training with synthetic data and labels from large-scale models, so I distilled and trained the 157GB weight DeepSeek V4 Fla… — @taroleo · 2026-09-19 · 模型原理 · 成本速度 · 内容审核
  • Jev is fast and cheap but is not very smart. No reasoning, no thinking. I don’t have benchmark but there are many questionable decisions in this Tetri… — @tdinh_me · 2026-09-18 · 游戏对战 · 模型原理 · 成本速度 · 评测对比
  • Analyzed a few thousand hours of agentic runs with @typesafeai Jev - turns out it's: • The best option I've tested at measuring progress and estimatin… — @hrishioa · 2026-09-21 · 编程开发 · 智能体 · 内容审核

📝 长文精选(Top 50)

  • The gains aren’t free: Jev can't generate text Comparing Jev vs LLMs side-by-side makes the trade-off clear Fun fact: replacing sequential computation… — @CompleteSkeptic · 2026-09-15 · 编程开发 · 成本速度 · 评测对比
  • 【Breaking News】ChatGPT Co-Inventor Unveils New AI Model "Jev" After 2 Years of Stealth Development ・20-200x faster than LLMs, 40-400x cheaper ・Output … — @ctgptlb · 2026-09-16 · 成本速度 · 评测对比
  • I hurriedly wrote a blog post. It's about Jev! The recently released "Jev" has become a huge topic of conversation, and in this post, I explain what s… — @nasuvit_z · 2026-09-17 · 编程开发 · 智能体
  • 🔥码住:如何最快使用Jev模型?输入只要 $0.04 / 1M tokens,比 LLM 最高便宜 444.6 倍、快 193.6 倍! Jev 已上 Vercel AI Gateway。状态进去,Boolean / Choice / Score + 概率直接出来,70–500ms 就能决定、路由、… — @NFT_Chen · 2026-09-17 · 成本速度
  • 最近Jev爆火,拿到 Jev API 还不知道怎么玩的,可以按这个清单抄。 1. jev-ultrafast Browser Use 做的高速浏览器 Agent。Jev 每一步只判断「做什么、点哪个元素」,要打字才叫小模型。Google Flights 搜一次航班大约 7 秒。 https:// g… — @Pluvio9yte · 2026-09-21 · 编程开发 · 智能体 · 成本速度
  • 昨天整理了 14 个 Jev 项目, @yangyi 开玩笑说:「让 Codex 一个 goal 跑个 Jev 导航站。」 我真的让 Codex 去跑了,现在导航站已经上线: https:// logicrw.github.io/awesome-jev-pr ojects/ … 而且增加到 130+… — @0xLogicrw · 2026-09-18 · 编程开发 · 工具资源
  • Yo! Brothers! Just from sleeping one night, the number of Jev-related GitHub repos has exploded!! I've compiled the ones with the most stars by genre!… — @studio_yebisu · 2026-09-18 · 编程开发 · 工具资源
  • As an LLM developer, this is the kind of application that really hits home, right? The end product is an LLM, not a classifier, so it shouldn't fall u… — @odashi_t · 2026-09-21 · 编程开发 · 内容审核
  • we've been exploring how to use Jev to build a better harness doing a webinar on it next week with @sydneyrunkle @huntlovell and Allie from @typesafea… — @hwchase17 · 2026-09-20
  • 想了解 Jev 的可以看小木老师这篇,从 0 到 1 讲得很清楚 看完想上手的,我给大家做了个「Jev 小白实战教练」Agent 筛交易信号、分析内容、做游戏 AI、操作电脑……把你想做的事告诉它,它帮你判断能不能做、需要什么工具、第一步怎么试 安装即用: https:// agent.creao.… — @Jason23818126 · 2026-09-20 · 智能体
  • It’s been an absolutely insane 72 hours since @TypeSafe launched its first public model, Jev ~18 months ago, @CompleteSkeptic , @EGafni , and @hackgoo… — @hardimanjames · 2026-09-19 · 成本速度
  • I was going to do a writeup of all the Jev stuff but someone already did and explained the thing I made yday :’) https:// sgnt.ai/p/jev/ tl;dr the dec… — @ekzhang1 · 2026-09-19 · 成本速度 · 内容审核
  • JEV is a cheat code for outbound. Deep lead qualification and outreach prioritization used to cost us a fortune. GPT and Claude analyzed prospects one… — @romanbuildsaas · 2026-09-18 · 编程开发 · 成本速度 · 金融交易
  • 🚀 TypeSafe AI's new model "Jev" is now publicly available without a waitlist! New users receive $5 worth of credits (approximately 120 million tokens … — @CryptoJPTrans · 2026-09-21 · 成本速度
  • Here's an example of Jev routing tools, skills, cli commands for Qwen 3.8 27B on @cerebras to format a Power BI report. Changes are almost instant; wo… — @kurtbuhler · 2026-09-18 · 编程开发 · 工具资源
  • With the launch of @typesafeai 's Jev, I have created a Claude Code plugin that scores how well you prompt your coding agent, and adds no latency at a… — @prateekkathal · 2026-09-18 · 编程开发 · 智能体 · 模型原理 · 成本速度 · 工具资源
  • Interesting to see how many people don't grasp Jev and ask on Twitter what that is. So here is the brief. Dimwits: 1. open chatgpt/claude 2. type "wha… — @tomhacks · 2026-09-18 · 成本速度 · 内容审核
  • 哈哈,Jev官方也审核通过了,一般申请一天后就过了 申请链接: https:// typesafe.ai 还没有的也不用着急,如下两个官方文档是最先值得看的: 1. Jev / System One 模型: https:// typesafe.ai/blog/introduci ng-system-… — @LufzzLiz · 2026-09-19 · 模型原理
  • For those who want to try using Jev from TypeSafe AI, I want you to be careful about this: It's waitlist-based, so there's a time lag from applying un… — @taishiyade · 2026-09-18
  • Found jev-mcp In tools like Cursor・Claude Code, everything is still handled by the LLM, but if we could insert Jev into the judgment part, it seems li… — @cu30rry_ · 2026-09-19 · 编程开发 · 成本速度
  • jev is a “System One” model from our friends at @typesafeai . it takes an input state and outputs probabilistic decisions. for example, you can pass i… — @michellechen · 2026-09-18 · 模型原理
  • Shoutout to @phanindra_ai for making this really awesome resource on Jev https:// jev-directory.netlify.app Everything Jev can do, with receipts. Runn… — @thisiskp_ · 2026-09-18 · 工具资源 · 评测对比
  • 目前能用上Jev的3个途径: 1. 官网加入候选名单 https:// typesafe.ai 2. openrouter https:// openrouter.ai/typesafe/jev-1 .13#playground … 3. Vercel 的 AI Gateway https:// v… — @akokoi1 · 2026-09-19
  • Jev @typesafeai 发布三天、刷屏 X 信息流! 再回看一个月前 TypeSafe AI CEO @CompleteSkeptic 在 @aiDotEngineer 主题演讲「Jev Creator: Why RLCD beats RLHF」,每一段论证几乎都是 Jev 产品规格的铺垫!… — @shao__meng · 2026-09-19 · 模型原理
  • jev costs $0.042 per million input tokens i’d try it on the ai call that only returns “billing” typesafe built it for structured decisions, with the p… — @dashboardlim · 2026-09-21 · 编程开发 · 模型原理 · 成本速度
  • Building harnesses is also changing thanks to Jev ( @typesafeai ) @sydneyrunkle nicely describes in her article how the routing model is changing, as … — @MAXIMISEART · 2026-09-18 · 内容审核
  • ลูกค้าทักมาว่า "เชื่อมบัญชีไม่ได้สามวันแล้ว" ระบบต้องรู้แค่ว่าเคสนี้ด่วนไหม แต่เรากลับต้องเรียกโมเดลภาษาอย่าง LLM ตัวเต็มมาตอบ Jev จาก TypeSafe AI คือ… — @vibecodingth · 2026-09-19
  • Jev performs really well in some evals for Executor Closest local equivalent is Gemma & vector search but it’s worse and bloats the local app Figuring… — @RhysSullivan · 2026-09-19 · 成本速度 · 评测对比
  • I switched Codex's model selection to Jev and started operating the router via TypeSafe. Posts I referenced: https:// x.com/su_nagomi/stat us/21015552… — @TaoRInne · 2026-09-21 · 编程开发
  • The bottleneck in coding agents isn't writing code anymore, it's the model deciding which of its own tools to look at. That's the quiet claim behind J… — @stretchcloud · 2026-09-21 · 编程开发 · 智能体
  • Did I just beat TypeSafe AI's Jev on calibration with an 11-second TF-IDF baseline ?! Here is what happened: my 151M model initially scored 26.10% on … — @heman10x · 2026-09-18 · 模型原理
  • Typesafe jev için aynı higgsfieldin uyguladığı social media hype yaratıp vurgun yapma stratejisini uyguluyor gibi hissediyorum iş ilanlarınıda açtılar… — @sercanyusuf_ · 2026-09-19 · 编程开发
  • Para @typesafeai y su increible JEV: Grosamente, piensa en esto como un MCP. No va además de lo que uses, va en conjunción. Lo que hace typesafe es ay… — @tebayoso · 2026-09-19 · 智能体
  • Why Your AI Still Lies to You (and the Radical Shift to "Smart Software") - Jev: Why RLCD beats RLHF — Diogo Almeida, TypeSafe AI The "Assistant" Trap… — @SungJinIn2 · 2026-09-18 · 模型原理 · 成本速度
  • i got jev + an llm playing doom, with a twist: try several futures, then keep playing from the best outcome. jev picks the moves. a custom harness use… — @toksdotdev · 2026-09-18 · 编程开发 · 游戏对战
  • Article 1: What is Jev + Comparison with GPT/Claude/All I spent 12+ hours this weekend reading everything about TypeSafe AI's new model. Here is what … — @parthjain_1 · 2026-09-21 · 评测对比
  • TypeSafe AI launched Jev, a model outputting probability decisions instead of text with completely free output tokens. The free output pricing caught … — @NFTMansa · 2026-09-18 · 模型原理 · 成本速度 · 内容审核
  • This Sunday, we’re building with Jev, @typesafeai ’s new System One model for fast, typed, probabilistic decisions. See real Jev architectures, then b… — @PromptQL · 2026-09-18 · 编程开发 · 智能体 · 模型原理 · 成本速度
  • Jev from TypeSafe AI is everywhere. Fast typed decisions. Not another chat model. What are you building with it? How are you wiring it into Claude Cod… — @_Zayn_Zahir · 2026-09-21 · 编程开发 · 智能体 · 成本速度
  • TypeSafe AI launched Jev this week, and my timeline is full of benchmarks. But most people are still testing it like a standard text model. Jev is a n… — @heman10x · 2026-09-18 · 编程开发 · 模型原理 · 评测对比
  • Were buzzing with use-cases at @arizeai - so many optimizations to be made. Trace every judgement Jev makes with one line of code! Instrumentation sup… — @ArizePhoenix · 2026-09-18 · 编程开发
  • Super early, few hours in. Cost effectivness combined with Codex has been interesting. Have to try to connect with some runbooks/AI-os workflow style … — @aravindbaskaran · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • Everyone talking about Jev from https:// typesafe.ai and built something similar to Jev for Chess. ChessJev is a System One model for chess and is not… — @gopalanj · 2026-09-18 · 编程开发 · 游戏对战 · 模型原理
  • إذا أول مرة تسمع بـ Jev، احفظ الاسم… لأن بعد أسبوعين التايم لاين عندك كله Jev 😂 Jev من TypeSafe AI موديل ما يسولف ولا يكتب؛ شغلته يقرر بسرعة. تعطيه خي… — @SAldwais · 2026-09-21
  • TypeSafe says Jev "never hallucinates." That claim is misleading. What they mean is that the output can't fall outside the schema you gave it. If the … — @JaceHall · 2026-09-19 · 编程开发 · 金融交易
  • Jev API 刚到手却不知道怎么用?下面这份清单就是给你准备的,直接抄就行。 一、接进日常工具 1. typesafe-mcp 刚拿到 key 就装这个。把 Jev 接到 Claude Code / Claude Desktop / Codex / Pi,随时问 Choice / Score / … — @CindyTaylo82399 · 2026-09-21 · 编程开发
  • A US frontier AI lab (TypeSafe AI) just announced a "breakthrough" System 1 architecture. The problem? An independent Indian developer built, publishe… — @Ramanean · 2026-09-19 · 编程开发 · 模型原理 · 工具资源
  • tengo un sistema automatizado personal donde automatizo mis sesiones de claude code/codex/cursor/fx y extraigo patrones (de ahí saco Skills y scripts)… — @tacosdedatos · 2026-09-19 · 编程开发 · 智能体
  • I made awesome-jev: about 50 open-source projects built on TypeSafe's Jev, sorted by stars. Browser and macOS agents, Android control, Claude Code plu… — @ckaraca · 2026-09-19 · 编程开发 · 智能体 · 工具资源
  • 緊急で社内Jev勉強会を開催したらアイデアがバンバン出て頭が柔らかくなった - LayerX エンジニアブログ https:// tech.layerx.co.jp/entry/2026/09/ 18/185816 … https:// b.hatena.ne.jp/entry/s/tech.l a… — @hatebu100 · 2026-09-18

🇨🇳 中文内容精选(Top 50)

  • 完全睡不着了....这个视频让你完全看懂JEV的恐怖能力 我实际演示了一下,用 JEV 和 DeepSeek 做同一个任务:从当天的海量实时新闻流里,为 15 个品牌快速匹配有没有适合借势的公关热点,并打上结构化意图分类标签 结果是:28 秒内,Jev 狂刷完了整整 428 — @SUOHA_AI · 2026-09-18
  • 💥炸了!开源本地版 Laya 响应速度吊打 Jev ,毫秒级决策快到飞起! 左边 Laya 是本地 421M 开源决策模型,右边 Jev 1.13.0 是云端 API。 同一盘贪吃蛇、同一套 typed decision,30 秒自由跑下来: 🔹Laya:分数 46、长度 52,决策 86.5 次/… — @NFT_Chen · 2026-09-20 · 编程开发 · 成本速度
  • 🔥速码!谁的 Codex 还没装上 Jev 批量必杀技?这波真能起飞 ! Jev 不写长文、不瞎规划,专门干一件事:选择、分类、打分。Agent 给上下文、自己干活。大规模数据清洗、自循环分类、长程任务里替你拍板,比把 Codex / Claude 当 judge 更便宜、更快、更能并发。 亮点先看… — @NFT_Chen · 2026-09-20 · 编程开发 · 智能体
  • 昨天出的 https:// typesafe.ai 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… — @coolish · 2026-09-17
  • Codex 接入 Jev 后,我的token额度消耗节省了90% 我实测输入一句: “游戏机我要买吗?” Jev 很快判断为 “购买咨询”,概率 97%,置信度 0.96,然后直接进入后续处理 它最适合干这些事,分类、筛选、路由、快速判断 Codex 则继续负责复杂推理。 接入也非常简单: 先去 h… — @goan999999 · 2026-09-20 · 编程开发 · 成本速度
  • Browser Use 接上 JEV 这速度太他妈吓人了吧?? 过去大家看 GPT、Claude 操控浏览器,体验其实并不好,因为真的太慢了,中间还经常卡死... 因为让大模型写一堆长篇大论去操作网页,本质上就是杀鸡用牛刀 在网页上点按钮,本质上就是一个离散的多选题,根本不需要 AI 会写诗 JEV… — @SUOHA_AI · 2026-09-17
  • 如何用上Jev: 1、官网申请加入waitlist,基本上申请当天能过。 🔗: https:// typesafe.ai 2、告诉你的Codex:安装npx skills add typesafe-ai/skills –skill typesafe-ai。 3、回到操作台,建立API Key。 4、… — @harrisonitsme · 2026-09-18 · 编程开发
  • 卧槽,居然真有比Codex内置computer use更快的操作电脑的方式 我尝试用Codex+Jev打造了一个加强版computer use,我称之为「Jev Use」。比内置的更快更丝滑,token消耗却差不多 下面是我用「添加Mac日历事件」做了一个对比视频,同样的内容,明显可以看到Jev版本… — @Saccc_c · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • Jev 太屌了!!!完全是模型的新方向+新范式。带来了太多不一样的东西。 用 OpenCLI 配合了下 Jev,爽到飞起。 以后其他大规模召回、AI推荐系统、实时交易,都可以用这个模型。 玩法太多了! — @jakevin7 · 2026-09-17
  • 直观感受下 seo 内链重构:jev pk opus5 Jev 在 45.1 秒内完成了全站重构,成功放置 584 个高相关性内链,并拒绝了 139 个强行凑数的无效页面,总耗费 $0.21。 作为对比,opus5 相同时间只处理了 21 个页面。 — @ai_xiaomu · 2026-09-19
  • 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… — @SUOHA_AI · 2026-09-18
  • 这也太夸张了吧!学到了!! 这拿来做竞品分析,简直有如神助!! 37 个品牌、724 支广告,40 秒拆完! 跑完一轮分析,成本0.09美元!! 做竞品分析的朋友,看到这个真的很难不心动。 Jev 批量分析竞品广告,系统拆解背后的思路,从开头怎么吸引人,到最后怎么促成行动,覆盖六个维度: — @yyyole · 2026-09-18
  • 牛逼!Jev 刚火没两天,已经有人搞了个开源的Jev🤣 @bespokelabsai 发布了一个开源版的Jev:Bespoke Nimble。 更离谱的是,整个项目是基于 Qwen3.5-9B ,只花了大约两天。 数据、模型权重、训练配方、推理代码全部公开,而且没有从 Jev 蒸馏,只拿 Jev 做… — @MaxForAI · 2026-09-19
  • 我用 @typesafeai JEV 模型做了一个实时 3D 场景生成器。 这个速度实在太夸张了! 从预制的几十上百种 3D 模型里,根据输入的文案,进行并发几百次的判断。 同时对这些素材的着色、光照、位置以及状态进行处理,一秒完成符合要求的室内场景搭建。 — @op7418 · 2026-09-20
  • 这几天 Jev 在 X 上突然爆火。 坦白说,我第一反应是没看懂:一个不能写文章、不能生成代码,只回答 Choice / Score / Yes-No 的模型,为什么会有这么多人兴奋? 我把能找到的增长相关 case 翻了一遍,才慢慢理解它可能在改变什么。 — @willlhhh · 2026-09-20
  • 🔥码住:如何最快使用Jev模型?输入只要 $0.04 / 1M tokens,比 LLM 最高便宜 444.6 倍、快 193.6 倍! Jev 已上 Vercel AI Gateway。状态进去,Boolean / Choice / Score + 概率直接出来,70–500ms 就能决定、路由、… — @NFT_Chen · 2026-09-17 · 成本速度
  • 💥我靠!用GPT-6 Astra把Jev运行结构可视化,直接把 LLM 和 JEV 的差别钉死了! 同一句:“网络一直连不上,请帮我处理。” LLM:一边在讲故事,一边在做判断。 🔹逐词元往下写 🔹前文决定下一词 🔹最后吐出一句完整话:「建议转给技术支持处理。」 🔹软件还得再解析这句话 — @NFT_Chen · 2026-09-18
  • 这几天全网都在跟风无脑吹 Jev,概念满天飞,但 99% 的人根本不知道怎么把它变成真金白银的生产力, @leaf_sanren 刚写的这篇长文给的8个案例都挺代表性的, 没扯啥 AGI 宏大叙事,开门见山直接甩出 8 个带源码仓库的落地案例,甚至极其硬核地把官方文档里藏着的致命雷区全扒了出来哈哈, — @AYi_AInotes · 2026-09-19
  • 💥实力比拼:Jev vs DeepSeek V4.1 Flash,瞬间秒懂Jev的恐怖能力! 500条真实电商客服工单,同一套分单任务: 🔸JEV:83秒清完,花了 $0.01,左边已经全部归档。 🔹DeepSeek:这轮停单时只做完 173条,花了 $0.06,右边还剩一大半,贵了大约5倍。 差距… — @NFT_Chen · 2026-09-19 · 评测对比
  • 停停停!现在要做的事是: Jev 模型橙皮书 Awesome -Jev GitHub 开源仓库 手把手带你入门 Jev 最值得推荐的 10 个 Jev 最佳案例 Jev 接入任意 Agent 教程 Jev 从入门到变现 文科生分享如何用 Jev 一个月赚到 100 万元 — @Jackywine · 2026-09-19 · 编程开发 · 智能体 · 工具资源
  • 💥牛了!将任意 LLM 秒变 Jev 模型实现本地化决策引擎! LLM2Jev 把 Jev 同一套 Choice / Score / Noul 搬到本地:现成 HuggingFace 因果模型,不调云、不买商业 API。 关键亮点 🔹运行时定义题型,直接拿带概率的类型化答案 🔹Prefill-onl… — @NFT_Chen · 2026-09-20 · 编程开发
  • 💥我靠!这是目前看到的给 Jev 模型找到的最狠的真实用例! Instant Compaction只删,不改写。 演示里 156k tokens 的 Claude Code 会话,Jev 逐条给 tool call 打分,瞬间压到 62k。 用户目标、精确报错、关键代码原文全在;过期 Read / … — @NFT_Chen · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • 终于看到了大语言模型之后的一个新的模型范式 Jev 是系统1模型,极速响应,价格超低,但不经过语言系统,不能说话,完全客观,也不需要语言层面的对齐。 潜力很大,也许是通向 AGI 的另一条路。 — @oran_ge · 2026-09-17
  • 有朋友用上 jev 了吗? 最佳实践的场景是什么? — @realchendahuang · 2026-09-20
  • Jev 🔥了 可是好多人连Jev是啥都不知道 没关系,看这一篇就够了! 包教包会,不会顺着网线去打他 @GeekCatX — @Lonely__MH · 2026-09-18
  • 最近 Jev 挺火,我拿它去 12306 查了趟动车票,看看这种浏览器 Agent 到底能不能干点日常的活。 用的是 Browser Use 开源的 Jev Ultrafast,一个主打速度的浏览器 Agent,可以自己操作网页,完成你交给它的任务。 我这次用 Pi + DeepSeek — @yanhua1010 · 2026-09-19 · 智能体 · 成本速度
  • 🔥推荐!让 Codex 搭配 Jev 成双王炸,浏览器操作速度提升 5–10 倍! 痛点很简单:Codex 自己点页面时,每一步都要读 DOM、推理、再调工具,来回一卡就是半天。 现在让 Jev 把「点、跳、切、滚」抢走,Codex 只负责想、输入、看图验收。 主要亮点: 🔹Jev 连续跑动作,Co… — @NFT_Chen · 2026-09-18 · 编程开发
  • 强烈建议大家去看一下 Jev 创始人 Diogo Almeida 的这个演讲,关于 Jev 的一切,他在这个视频中早就预言。 Jev 一开始就是盯着自动化去的,Diogo Almeida 认为现在的 LLM 陷入了一个史诗级大弯路,就是 RLHF。 RLHF — @threeaus · 2026-09-20
  • 太爽了!终于找到了JEV的最强应用场景之一--全自动化 Browser use 我用JEV全自动填完 TypeSafe AI 的申请问卷,我只给它丢了一个完全陌生的网址,让它全自动帮我完整填完:38 秒一次性刷完 16 道题,全程没停顿、零人工干预 具体的逻辑是: 我看到 @typesafeai 文… — @SUOHA_AI · 2026-09-20
  • 最近Jev爆火,拿到 Jev API 还不知道怎么玩的,可以按这个清单抄。 1. jev-ultrafast Browser Use 做的高速浏览器 Agent。Jev 每一步只判断「做什么、点哪个元素」,要打字才叫小模型。Google Flights 搜一次航班大约 7 秒。 https:// g… — @Pluvio9yte · 2026-09-21 · 编程开发 · 智能体 · 成本速度
  • Langchain 的速度太快了,已经利用 Jev 增强 harness。 跟昨天讨论的一致,Jev 非常适合在既定 harness 中承担分析分类工作,例如 agent 路由,模型路由等螺丝钉工作。 — @dongxi_nlp · 2026-09-18 · 智能体
  • 刚搜了下,jev 马上就有一大堆替代 Semif 可以看看,用Qwen3.5-4B,无需训练,就可以做到一样的准确率,而且更便宜。macbook 本地随便部署。 LLM 已经是训练的很好的东西,真没必要再自造新的。 https:// github.com/TheoLeeCJ/SemIf https:… — @9hills · 2026-09-20 · 编程开发 · 模型原理 · 评测对比
  • 最应该把 JEV 用对的地方:闲鱼⬇️ — @miles_mazy · 2026-09-20
  • 这两天最火的新构架大模型 Jev,官方Waitlist填写地址 Vercel AI Gateway已经可以调用,明天抽空试试。 两个地址见评论区,建议收藏。 很多朋友在聊,但具体应用场景还需要挖掘。 — @vista8 · 2026-09-17
  • Jev 太离谱了。 只用 40 秒,它就拆解了 37 个品牌正在投放的 724 条广告。 每条广告的钩子、形式、Offer、CTA、用户认知阶段,以及广告与落地页不匹配的问题,全都分析了出来。 Token 成本仅 0.09 美元。即将接入 StealAds + MCP。 — @FinanceYF5 · 2026-09-20 · 成本速度
  • 卧槽! 我把 Jev 玩到了 Grok Bot 里 跑 X 的时间线浏览, 看有没有贴子适合自己回复下, 这可是大大的省时间了 整个流程十分简单, 照着文章操作几分钟就能用起来 — @0xlangeai · 2026-09-20
  • 🔥 卧槽了,jev 简直太适合 codex 用户了,强烈建议 codex 用户去尝试下 jev 因为 jev 非常适合需要快速进行复杂决策和评估的场景,所以和 codex 的 computer use 功能简直是绝配。 直接把 jev 用作判官,指导 codex 快速操作你的电脑,既准确又节省 to… — @alin_zone · 2026-09-18 · 编程开发 · 智能体 · 成本速度
  • 最近应该都被Jev 刷屏了。 但是该怎么玩呢? 一个视频教你在 WorkBuddy 中接入 Jev。 接入完该拿它来干什么呢? 它不会聊天、不会写代码、一个字都吐不出来。 只回答三种题:选哪个、打几分、是或否。 虽然大概率给的答案是正确的,但不代表不会出错。 所以最适合它的,不是「替代大模型」。 — @aehyok · 2026-09-19
  • 昨天整理了 14 个 Jev 项目, @yangyi 开玩笑说:「让 Codex 一个 goal 跑个 Jev 导航站。」 我真的让 Codex 去跑了,现在导航站已经上线: https:// logicrw.github.io/awesome-jev-pr ojects/ … 而且增加到 130+… — @0xLogicrw · 2026-09-18 · 编程开发 · 工具资源
  • Jev模型的Waitlist提交后没多久就通过了。 也用Codex调用成功,但还没想到具体应用场景。 「Jev 是一个非常强的、高性价比的决策层」 让 AI 整理一份资料,要点总结下: 1. Jev 是一个专为软件自动化决策设计的模型,不能聊天、不能写文章,只做决策。 — @vista8 · 2026-09-18 · 编程开发
  • 别管 Jev 到底有啥用,能赚钱就是好东西! — @miles_mazy · 2026-09-19
  • Jev 爆火没几天呢,它的开源对手 Laya 就出现了。 两者都不生成长文本,而是直接输出选择、评分和概率,用来完成路由、审核、风险判断等快思考、快决策任务。 不同的是,Jev 闭源研发约两年,只提供 API;Laya 约 3~4 亿参数,代码和权重全部开放,可以在本地运行和微调。 在 GitHub — @Yangtze_Seventh · 2026-09-20 · 编程开发
  • Jev 发布没几天,开源社区已经开始疯狂复刻了🔥 最值得推荐的五个模型: 1、Laya 421M:原生决策模型,支持 Mac 2、Decider-2B:最像 Jev,基于 Qwen3.5 3、NanoJev 0.6B:专门的 Decision Head 4、Reflex:Qwen3.5 + Dire… — @xiaomovps · 2026-09-18 · 编程开发
  • 用AI一定要试试Jev :跳过聊天,直接输出带置信度的结构化决策。特定任务快193倍、便宜445倍;开源插件能快速压缩上下文,长任务告别越跑越“失忆”。 Jev上手很简单: 1、去 https:// TypeSafe.ai 申请资格 2、让 Codex 安装 TypeSafe Skill 3、在后台… — @KKaWSB · 2026-09-18 · 编程开发
  • Jev是一个决策小模型,但是从第一性原理来看 Jev真的不如离线设备跑一个 Qwen 3.8 Flash Next 或者离线小模型更香,这些离线模型除了可以做决策以外,还可以干很多其他的事情 当然我也相信社区很快就可以开发完全离线的 Jev 模型出来 那样很多自动化用电脑的方式就可以超快、隐私而且成… — @manateelazycat · 2026-09-17
  • 现在,TypeSafe 的 JEV 模型已经全量开放了,不需要申请等待列表。 而且每个人有 5 美元的免费额度,这模型非常便宜,5 美元够你用好久了!可以去试试我这个案例了 — @op7418 · 2026-09-21
  • Jev 不写邮件,它只做一件更狠的事:判断这封邮件该不该发给这个人! 700 个高意向线索 × 个性化文案 40 秒打完:预测回复、给置信度、抓人货错配 成本只有 $0.09 SDR 花两小时想的事,它当一次函数调用。 大模型负责写,Jev 负责判。 高置信自动发,低置信丢给真人。 GTM — @NFT_Chen · 2026-09-18
  • 卧槽, 我让 Jev 新模型来分析下 Claude 封号问题,它居然全都判断正确了…… 这个决策太强了啊!! Jev 这不是还能做意图识别? Jev 这是从 OpenAI、A/ 中杀出一条血路…… — @xmglab · 2026-09-21
  • 这是叶总把 JEV 用到我已有的开发的场景里面,了解的过程中,索性就出一篇汇总文章 , 很值得看一看 — @0xlangeai · 2026-09-19
  • 在闲鱼发了 jev 代开服务,已有第一单 — @EthanWang07X · 2026-09-19

数据说明

  • 采集方式:X 站内搜索(Jev / TypeSafe / typesafeai / filter:videos / lang:zh),滚动加载后解析,按发布时间去重。
  • 收录口径:提及 Jev / TypeSafe AI 的动态,剔除明显无关与代币类噪音。
  • 类型判定(互斥):含视频 → 演示视频;无视频且正文 ≥ 200 字符 → 长文;其余归入讨论。三者相加等于总数。
  • 完整 982 条结构化数据见 data/posts.json,含作者、时间、原文、标签、原帖链接。

关于作者

谁是专家 — 小红书 · 微信公众号 · X 同名

版权与来源

所有文字、图片、视频版权归原作者所有。本项目仅做信息聚合与索引,全部内容以链接形式指向 X 原帖。 如原作者希望移除其内容,请提 Issue,我们会第一时间处理。


出品:谁是专家 | @dboy_yi2025

About

JEV HUB · X 上关于 TypeSafe AI「系统一模型」Jev 的长文与演示视频聚合(保留原链与作者)| 谁是专家 出品

Topics

Resources

Stars

25 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages