shipwithjev

Blog / Questions / FIG. 126

Who Uses Jev? The Builder Census

Who uses Jev? A census of cataloged builds by category and source, from solo GitHub tinkerers to Metaview, Vercel, and LangChain integrations.

Mostly independent developers shipping open-source tools, plus a fast-growing layer of companies wiring it into agent platforms and frameworks. That's the answer from counting the catalog rather than guessing: this site tracks what people publicly build with Jev, TypeSafe AI's decision model, and the census below is those entries sorted.

One scope note: this counts public builds. TypeSafe hasn't published customer numbers we can cite, and private production use is, by definition, invisible to us.

The census by category

At the snapshot we counted, the catalog held 551 entries. By category:

  • Tools and apps: 213. Linters, CLI helpers, browser extensions, inbox filters. The long tail of "I had a small decision problem."
  • Agents and browsers: 79. Step-by-step action selection for browser and desktop agents.
  • Research and data: 72. Labeling piles of text, benchmarks, studies.
  • Games and real-time: 66. Doom, Tetris, Pokemon, and everything that proves latency in public.
  • Content and growth: 54. Post scoring, slop filters, ad teardowns.
  • Triage and routing: 50. Email, tickets, PR labels, moderation.
  • Trading and markets: 10. Small, loud, and heavily caveated.
  • Robotics and devices: 7. The newest frontier, mostly simulations.

By source, 278 entries are GitHub repos, 183 are X posts, 46 are write-ups, and 26 came from Reddit. The builder base skews technical and open-source: people who ship code first and tweet the receipt second. The story of how that first wave arrived belongs to the launch-week page; what each category actually does lives on the use-cases page.

The companies in the catalog

Company adoption shows up as integrations more than announcements. Metaview reported shipping Jev into every agent on its platform, with sourcing searches going from minutes to seconds at the same accuracy, as reported (build). Vercel's open agent framework, eve, uses Jev as its default evaluate model (build). LangChain, LiteLLM, and Composio all have cataloged integrations, and Every's Mike Taylor ran 1,709 editorial judgments for under a cent, as reported (build).

The honest read: frameworks integrating a model is a signal of developer demand, not proof of production scale. Watch which integrations are still maintained in a quarter.

Frequently asked questions

Who uses Jev the most?

By cataloged public builds, independent developers publishing open-source tools, with tools-and-apps the largest category at 213 of 551 entries in our snapshot.

Which companies use Jev?

Cataloged examples include Metaview's agent platform, Vercel's eve framework, and integrations from LangChain, LiteLLM, and Composio. TypeSafe AI hasn't published a customer list we can cite.

What do most Jev users build?

Small decision tools, agent step-selection, research labeling, and real-time games lead the count; the use-cases page breaks down each category with receipts.

How do I get my build counted?

Submit it with a public link and any numbers you measured; the census updates as the catalog grows.

Numbers throughout are as reported by the build authors, not verified by shipwithjev. Code-shaped examples are pseudocode; the official docs live at docs.typesafe.ai.