As of: 2026-11-19 22:11 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/p/x-airesearch12-2101311992984199580/ # JevBench results Florian S (@airesearch12), X, 2026-09-19. Source: https://x.com/airesearch12/status/2101311992984199580 ## Why it matters Community benchmark for Jev-class models (212k views). ## Summary Community 'JevBench' (benchmarkheaven.com): original Jev leads at 75.3 with SemIf second at 74.6, per its author; 'Jev still in the lead, but it's close.' Unofficial benchmark. _Metrics read 2026-09-29 via api.fxtwitter.com: 212,393 views, 1,007 likes._ ## Archived text > JevBench results are in. > > Jev still in the lead, but it's close. > > > Quoting @airesearch12: Introducing JevBench. > > > > The first benchmark for Jev class models. > > > > Original Jev by @typesafeai in the lead at 75.3. > > SemIf #2 at 74.6. > > > > All results at https://t.co/hFQ5fX4JEb https://t.co/nXYfCrZnwO Archived 2026-09-29 via syndication. Counts: 1,007 likes, 59 replies (at fetch time) Media: https://pbs.twimg.com/media/HSlcDnPXkAAHTH1.jpg ## Cited in - 2026-09-15: [TypeSafe AI releases Jev, a 'System One' decision model that returns typed probabilities instead of text](https://postcutoff.com/e/2026-09-15-typesafe-jev-system-one-model/)