shipwithjev
shipwithjev is an independent, community-run directory of things people have built on Jev — TypeSafe AI's System One decision model. It's not run by TypeSafe or by us: it's a hand-sorted catalog at shipwithjev.com, where each build links to the X post, repo or site behind it, and many carry the cost and latency their author reported. Here's what's inside, and how to run the same kind of Jev decision yourself.
Jev returns a typed decision — a choice, a score, or a calibrated yes/no (noul) — in a single non-autoregressive pass, so it slots into anything that needs a fast, bounded judgment instead of generated prose. That's why a catalog like shipwithjev filled up so fast: once a decision costs well under a cent and lands in tens of milliseconds, people wire it into agents, routers, judges, games and data pipelines. The directory is the easiest way to browse what that looks like in the wild.
What's in the catalog
As catalogued, shipwithjev lists 620 builds across 16 pages, grouped into eight categories and drawn from X posts, GitHub repos, Reddit threads, videos and sites. The spread shows where Jev is actually being used:
| Category | Builds |
|---|---|
| Tools & apps | 234 |
| Agents & browsers | 89 |
| Research & data | 85 |
| Games & real time | 77 |
| Content & growth | 58 |
| Triage & routing | 55 |
| Trading & markets | 12 |
| Robotics & devices | 10 |
The free tools
Alongside the directory, shipwithjev runs a set of small free tools: about half call Jev and show you what each run cost (a tone checker, a résumé-to-job-description matcher and the like), and the rest run entirely in your browser (a CSV cleaner, an LLM cost calculator). They're a quick way to feel what a typed decision is before you write any code — and they make the per-decision cost concrete.
Get your build listed
shipwithjev takes submissions, so if you ship something on Jev you can add it to the catalog; the site also sells a weekly sponsor slot. Because it's community-run and not affiliated with TypeSafe, treat the reported cost and latency figures as each author's own numbers, not official benchmarks — and if you want calibrated, pinned numbers you can trust a threshold against, run your own decision on the hosted endpoint.
Want to build the thing that lands in the catalog next? The fastest start is to run a real Jev decision in the browser here, then drop a jv_live_ key into your code — the same hosted endpoint every build on the list is calling.
FAQ
What is shipwithjev?
An independent, community-run catalog at shipwithjev.com of projects built on Jev, TypeSafe AI's System One decision model. It lists 600+ hand-sorted builds grouped by category, each linking to the post, repo or site behind it, with many carrying the author's reported cost and latency. It is not run by TypeSafe or by us.
Is shipwithjev made by TypeSafe AI?
No. The site states it is not affiliated with TypeSafe AI — it's a community directory that happens to focus entirely on Jev. We're not affiliated with it either; this page just explains what it is and links to it.
How many builds does shipwithjev list?
620 builds across 16 pages as catalogued, grouped into eight categories — Tools & apps (234) is the largest, followed by Agents & browsers (89) and Research & data (85). The listings are drawn from X posts, GitHub repos, Reddit threads, videos and sites.
Can I add my own Jev build?
Yes — shipwithjev accepts submissions, and it also offers a paid weekly sponsor slot. If you want to build something to submit, run a decision on the hosted Jev endpoint first so you know your cost and latency are the calibrated, pinned ones.
See also: Made with Jev · Awesome Jev lists · JevBench leaderboards · Try Jev in the browser
Build the next one on the list
Run a real, calibrated Jev decision free in the browser, then grab a jv_live_ key — the same hosted endpoint every build in the catalog is calling.