TypeSafe AI opened early access to Jev and called it its first public System One model. System One is a class of models for fast, structured decisions that software can use directly.
Jev takes unstructured data and returns probabilistic decisions of a specified type. TypeSafe says Jev’s response time is 70–500 ms.
Claim check:
- TypeSafe AI opened early access to Jev and called it its first public System One model. (confirmed by the publication itself: evidence; «Our first public model is Jev , available today in early access.»)
- TypeSafe AI describes System One as a new class of models for fast, structured decisions that software can use directly. (confirmed by the publication itself: evidence; «a new class of frontier models built to make fast, structured decisions that software can use directly.»)
- Jev takes unstructured data and returns probabilistic decisions of a specified type. (confirmed by the publication itself: evidence; «Think of Jev as a frontier-intelligence function call: unstructured state in, typed probabilistic decisions out.»)
- TypeSafe says Jev’s response time is 70–500 ms. (confirmed by the publication itself: evidence; «End-to-end response time is 70ms-500ms for TypeSafe.»)
Publications:
- https://techcrunch.com/2026/09/18/a-new-kind-of-ai-model-from-a-chatgpt-inventor-is-thrilling-developers
- https://registerspill.thorstenball.com/p/joy-and-curiosity-100
- https://producthunt.com/products/jev-2
- https://kdnuggets.com/what-everyone-is-getting-wrong-about-typesafe-ais-jev
- https://thenewstack.io/typesafe-jev-system-one
- https://simonwillison.net/2026/Sep/21/jev
- https://reddit.com/r/LocalLLaMA/comments/1wijo3e/i_literally_built_the_jev_architecture_one_year
- https://nobodywho.ai/posts/jev-in-25-lines
- https://alexmolas.com/2026/09/23/jev-cant-be-calibrated.html
Primary sources:
- https://typesafe.ai/blog/introducing-system-one-models-and-jev
- https://typesafe.ai/blog/bitterest-lesson
- https://einride.tech/press/einride-and-lidl-launch-first-autonomous-cab-less-truck-on-german-public-road
- https://github.com/chaseleantj/desktop-habitats
- https://github.com/saragordic/window-sweaters
- https://github.com/Futureppo/typesafe_register
- https://console.typesafe.ai/playground
- https://huggingface.co/facebook/bart-large-mnli
- https://github.com/browser-use/jev-ultrafast
- https://vercel.com/changelog/ai-gateway-now-supports-typesafe-clients-and-http-api-for-jev
- https://latent.space/p/jev
- https://news.ycombinator.com/item?id=49717558#49718407
- https://github.com/kyle-pena-nlp/jevchat
- https://github.com/f/jev-leftpad
- https://gist.github.com/cablehead/bdf9ad946ceb26d9008976e49c9bfbbb
- https://github.com/jaredpalmer/kev
- https://arxiv.org/abs/2503.23303
- https://huggingface.co/DeepMostInnovations/sales-conversion-model-reinf-learning
- https://huggingface.co/datasets/DeepMostInnovations/saas-sales-conversations
- https://arxiv.org/abs/2510.01237
- https://github.com/sandeco/pix-golpe
- https://github.com/lyramakesmusic/jevbot
- https://openjev.com/
- https://github.com/ekzhang/openjev-sglang
- https://github.com/razorback16/openjev
- https://tenuo.ai/blog/jev-scoped-authority
- https://docs.typesafe.ai/introduction/machine-learning-primer#rlcd-and-calibrated-decisions
- https://www.distillabs.ai/blog/jev-or-a-fine-tuned-small-model-we-built-a-pipeline-with-both-to-see-the-real-difference
score 50.2 out of 100 · kind: announcement · update 15