JEV · CURATED DISCOVERY
Explore TypeSafe Jev in one place, from launch posts to real demos. Find use cases by industry, see original video frames, and open every source post from X.
| 4814 curated posts |
2483 original videos |
25 categories |
0.72 minimum confidence |
From finance and gaming to security review. Every listed post is checked against its X source.
|
社区实践 813 showcases · 336 videos |
工作原理 504 showcases · 221 videos |
开发者工具 486 showcases · 247 videos |
|
接入动态 372 showcases · 155 videos |
产品发布 370 showcases · 176 videos |
智能体 359 showcases · 206 videos |
|
游戏 335 showcases · 255 videos |
开发资源 307 showcases · 121 videos |
金融 238 showcases · 133 videos |
|
自动化工作流 203 showcases · 136 videos |
营销 133 showcases · 100 videos |
内容创作 108 showcases · 81 videos |
|
知识与搜索 101 showcases · 54 videos |
数据分析 98 showcases · 54 videos |
科研 68 showcases · 25 videos |
|
网络安全 56 showcases · 18 videos |
客户支持 48 showcases · 30 videos |
机器人 41 showcases · 32 videos |
|
生产力 35 showcases · 25 videos |
软件开发 28 showcases · 13 videos |
企业运营 27 showcases · 14 videos |
|
医疗 22 showcases · 16 videos |
教育 22 showcases · 11 videos |
法律合规 20 showcases · 6 videos |
|
电商 20 showcases · 18 videos |
Recently added Jev posts, ordered by source publish date.
|
01 
The author uses Grok to hunt viral X posts, JEV to decide what deserves attention, and only shows up to approve. 0 views · 💬 0 🔁 0 ♡ 0 📊 0 kiosa · @thegreatest_sv |
02 
Jev achieved 64/72 human labels on a harder LLM-judge dataset, while GPT-OSS-120B scored 72/72. Jev's median latency is 0.20s vs 1.44s, with an estimated cost of $0.020 per 1,000 judgments. Retrospective analysis shows routing uncertain judgments to a second model corrected observed misses. 27 views · 💬 0 🔁 0 ♡ 0 📊 27 MLflow · @MLflow |
|
03 
EM360Tech explains how Jev AI makes structured decisions, how it differs from an LLM, and where decision models could fit into enterprise AI. 8 views · 💬 0 🔁 0 ♡ 0 📊 8 EM360Tech · @EM360Tech |
04 
Jev, from TypeSafe, is a decision model that picks a category, answers yes/no, or scores against a scale. This post provides a hands-on explanation of what a decision model is, why it differs from a chat model, and how to use it to build your own tools. 10 views · 💬 1 🔁 0 ♡ 0 📊 10 Charles Shen · @agenteer |
Ranked by views among the posts in this curated index.
|
01 
A user shares an ideal use case for Jev: instant compaction. By scoring every tool call and dropping irrelevant content, Jev can make compaction instant instead of relying on summarization prompts. 3.6M views · 💬 404 🔁 967 ♡ 11K 📊 3.6M tamara · @tamarajtran |
02 
Gregor Zunic showcases an open-source browser agent combining Browser Use and Jev, finding flights in 7 seconds for $0.0039, using a new action space per step, DOM state space, and small LLM fallback for typing. 3.0M views · 💬 264 🔁 938 ♡ 8.8K 📊 3.0M Gregor Zunic · @gregpr07 |
|
03 
Developer kimura512 built a site using Jev that lets users freely input A and B to quickly judge whether A is B, such as "Is a banana a snack?", with a link to try it. 3.0M views · 💬 58 🔁 1.1K ♡ 1.9K 📊 3.0M kimura512@開発者 · @Bioinfo_Kimura |
04 
The author shares a 45-second TL;DR video on Jev, finding the core idea simple but the original video hard to follow. 1.6M views · 💬 220 🔁 949 ♡ 9.4K 📊 1.6M Matija Sosic · @MatijaSosic |
Start with verified Jev posts.
|
01
The author uses Grok to hunt viral X posts, JEV to decide what deserves attention, and only shows up to approve. 0 views · 💬 0 🔁 0 ♡ 0 📊 0 kiosa · @thegreatest_sv |
02
Jev, from TypeSafe, is a decision model that picks a category, answers yes/no, or scores against a scale. This post provides a hands-on explanation of what a decision model is, why it differs from a chat model, and how to use it to build your own tools. 10 views · 💬 1 🔁 0 ♡ 0 📊 10 Charles Shen · @agenteer |
|
03 
The author had Jev analyze 221 top posts from 7 creators in 51 seconds for just $0.032, making 1,547 decisions. Findings: how-to hooks save 3.2x the median, real numbers as proof 2.1x, and text-only posts only 0.5x. 4 views · 💬 0 🔁 0 ♡ 0 📊 4 Mert Durmazer · @mertdurmazer |
04 
In this tweet, 8Bit🦞 discusses an interesting take — if certain assumptions hold, ore's mining model could do much more than expected. He plans to experiment live using his new @ore computer miner, combining @typesafeai Jev with his own clef model to see if they can build something powerful. 166 views · 💬 0 🔁 1 ♡ 1 📊 166 8Bit🦞 · @0rdlibrary |
Source dataset: work/x-import/review.json
Titles, summaries, and categories are editorial. Posts, images, and videos belong to their original authors and platforms.