ToolsSeptember 16, 2026via Vercel Blog
TypeSafe AI's Jev now available on AI Gateway
Why it matters
Jev replaces token-by-token LLM generation with parallel probabilistic evaluation for structured decisions. Practitioners building agents or guardrails can now route uncertain cases to review while automating high-confidence paths—a meaningful efficiency gain for production workflows.
Key signals
- Jev available on Vercel AI Gateway as of Sep 16, 2026
- 193.6x faster and 444.6x cheaper than LLMs on workflow evaluations (TypeSafe claim)
- Returns typed Choice, Score, and Boolean answers with confidence/probability scores
- Integrated into AI SDK 7.0.105+ via experimental evaluate API
- Supports zero data retention and no training on user data
- Use cases: agent subagent selection, guardrail enforcement, scoring urgency/risk, output verification
- Jev available on Vercel AI Gateway (live as of Sep 16, 2026)
- 193.6x faster and 444.6x cheaper than LLMs on workflow evaluations (TypeSafe reported)
- Returns typed answers (Choice, Score, Boolean) with confidence/probability scores in parallel, not sequential token generation
- Use cases: agent decision routing, guardrail enforcement, output verification, priority/urgency scoring
- Supports Zero Data Retention and No Training modes per request
- Appears in logs, custom reporting, and counts toward budget tracking
The hook
193.6x faster, 444.6x cheaper. TypeSafe's Jev decision model is now live on Vercel's AI Gateway—typed answers without the text generation tax.
is now available on AI Gateway.Jev from TypeSafe AI
Jev is a probabilistic decision model for software: state goes in, typed Choice, Score, and Boolean answers come out.
Regular language models generate text one token at a time, which the application then parses and validates. Jev evaluates all decl…