FULL STORY
Typesafe AI's Jev: from benchmarks to Vercel adoption
Developers benchmarked Typesafe AI's new model Jev, finding it faster, cheaper and higher quality than Gemini 2.5 Flash Lite; Vercel then confirmed it will use Jev for fx auto's security checks, an ~18x speedup.
2026-09-16 ~ 2026-09-17 · 2 episodes · 6 posts
Episode 1 · Typesafe AI's Jev Beats Gemini 2.5 Flash Lite on Speed, Cost and Accuracy (2026-09-16, 3 posts)
Independent tests show Typesafe AI's new classification model Jev saturates an existing Gemini 2.5 Flash Lite benchmark at 100% accuracy, running about 6x faster while cutting costs by 98% and gaining 12 accuracy points in another evaluation.
- TypeSafe Jev early-access test: 5.7x faster responses, 98% lower cost, +12pp accuracy — jon_reed · 2026-09-16
- Typesafe AI's Jev saturates classifier eval and runs 6x faster than Gemini 2.5 Flash Lite — cramforce · 2026-09-17
- Typesafe AI's Jev saturates Vercel classifier eval at 100% accuracy, 6x faster than Gemini — schwentker · 2026-09-17
Episode 2 · Vercel fx to adopt Jev safety reviewer, up to 18x faster (2026-09-17, 3 posts)
Vercel CEO Guillermo Rauch confirmed that fx's default auto-mode safety reviewer will switch to typesafeai's Jev, which benchmarks show is roughly 5-18x faster and more accurate than the current GPT-5.6-Luna classifier.
- fx's safety classifier benchmarked: ~5-18x faster and more accurate than GPT-5.6-Luna — cramforce · 2026-09-17
- Vercel's fx may switch safety reviewer to Jev: 5-18x faster than GPT Luna — thesaraharminta · 2026-09-17
- Rauchh Hypes typesafeai: fx Safety Reviewer Runs on GPT Luna, 18x Faster p95 (Unverified) — schwentker · 2026-09-17