Jev with Vercel

Two ways in

Note: When using AI SDK direct, the key is TYPESAFE_AI_API_KEY, not TYPESAFE_API_KEY.

Endpoints: POST /typesafe/v1/systemone and GET /typesafe/v1/models. Auth: Gateway key or OIDC.

An experimental_decide call

One short call with one Choice question and one boolean question. Boolean maps to Noul.

import { experimental_decide } from 'ai';
import { typeSafeAi } from '@ai-sdk/typesafe-ai';

const result = await experimental_decide({
  model: typeSafeAi.decisionModel('jev-latest'),
  state: "Hi, I've been waiting 3 days for my refund. Still nothing.",
  questions: {
    department: {
      type: 'choice',
      instructions: 'Which team should handle this',
      criteria: {
        billing: 'Payment or subscription issues',
        technical: 'Integration or API problems',
        support: 'General help'
      }
    },
    is_urgent: {
      type: 'boolean',
      instructions: 'The message conveys urgency'
    }
  }
});

console.log(result.answers.department.choice); // "billing"
console.log(result.answers.is_urgent.probability); // P(true), 0 to 1

Confidence for Choice and Score is at result.providerMetadata.typesafe.confidence[questionId].

maxRetries defaults to 2 and covers 429 and 529.

Source: TypeSafe AI provider docs.

Note: experimental_evaluate is a deprecated alias of experimental_decide. Pong and the mahjong coach still use the old name.

Jev or a chat model behind the same call

AI SDK also accepts OpenAI, Anthropic and Google decision models.

Jev and OpenAI Decisions return native probability distributions.

The Anthropic and Google adapters return prompted P(true) estimates for Boolean questions, and no distributions for Choice or Score.

More at /compare/jev-vs-structured-outputs and /compare/jev-vs-openai-decisions.

Price, BYOK and data handling

Same list price, with no markup.

BYOK with a TypeSafe key bills TypeSafe directly.

Vercel's ZDR table lists TypeSafe AI with zero data retention ✓ and no prompt training ✓. Source: Vercel ZDR.

ZDR enforcement is Pro/Enterprise only: per request at no extra cost, or team-wide at $0.10 per 1,000 requests.

TypeSafe's own policy: no training on customer data, and ZDR for enterprise. Source: TypeSafe legal.

Decision fallbacks

A Gateway extension that reruns an uncertain answer on another model. Source: Vercel decision fallbacks.

A triggered request bills both stages.

Response headers x-ai-gateway-decision-fallback-* tell you when a fallback fired.

Limits on Vercel

• The API is experimental.
• Gateway lists 32,000-token context; TypeSafe's limit is 64k total.
• Jev takes text input only, even though AI SDK state can carry files for other models.

See Jev limitations for the model's own limits.

Reading

Guide·youtube.com/@GregIsenberg

Jev is HERE. How to use it

Greg Isenberg and Ryan Vogel on the Startup Ideas Pod: what Jev is, the businesses it unlocks, and how to call it today through the Vercel AI Gateway.

GREG ISENBERG

@gregisenberg

Jev is HERE and this is the CLEAREST explanation of what it is and what NEW businesses it unlocks. (and at the end I'll tell you how to get Jev even if you're on the waitlist) WHAT IT IS You know how you open your inbox and have to decide what's junk, what needs a reply, and what can wait? Jev does that part. It looks at each thing and says "this is junk, I'm 94% sure." It doesn't write anything back to you. It just sorts. 1,700 emails for 18 cents, instantly. That sounds kinda trivial but the important part WHAT IT UNLOCKS My explanation of Jev sounds small until you realize HOW MANY jobs are exactly this. Someone reading a stack of applications. Someone deciding which support ticket goes to which team. Someone looking at inbound and deciding who's worth calling back. A few ideas on what it unlocks: 1/ Instant quotes that are actually instant. Every quote form on the internet says "we'll email you by end of day." Build the version that answers in under a second, for roofers, movers, insurance, legal intake. 2/ Lead scoring as a product. Every agency and service business has a contact form full of junk. Score every submission and send the real ones straight to the owner's phone. 3/ Support triage for companies with no support team. The ticket gets classified and routed before anyone opens it. 4/ Clipping tools. Pass in a transcript, get the best moments scored in three seconds. Every clipping product just got a cheaper engine. 5/ Application piles. Grants, permits, insurance claims, job apps, loan docs. Someone reads that stack one item at a time today. 6/ Marketplace matching. Someone types what they need and gets matched to the right local business instantly instead of waiting for callbacks. 7/ Browser agents that actually move FAST. That makes bulk browser work practical: pulling quotes from five carriers, filing the same form for 200 clients, checking supplier inventory in real time etc. TLDR; find an expensive queue and put Jev at the front of it. HOW TO GET IT I didn't realize you can skip the waitlist because Jev is live on the Vercel AI Gateway right now, so you can start calling it today. In this episode, we share how. Episode now live on @startupideaspod (thanks to @ryanvogel for coming on and spilling the sauce today) Watch: https://www.youtube.com/watch?v=4mTLpuQpB80 Jev is a big deal because this is a whole new way to do AI Really cool Happy Jev day.

Guide·x.com

What Jev is, and the businesses it unlocks

Greg Isenberg: find an expensive queue and put Jev at the front of it. Seven ideas, from instant quotes to lead scoring, plus a Startup Ideas Pod episode with Ryan Vogel.

External link: Vercel Jev integrations.

Common questions

How do I use Jev with the Vercel AI SDK?

Call experimental_decide with typeSafeAi.decisionModel('jev-latest'), or with the string 'typesafe-ai/jev' through AI Gateway. The API is experimental.

What is typesafe-ai/jev?

Jev's model ID on Vercel AI Gateway.

Is Jev more expensive on Vercel?

No. It's listed at $0.042 per million input tokens with output free, and Gateway adds no markup.

Does Vercel AI Gateway keep my Jev data?

Vercel lists TypeSafe with zero data retention and no prompt training. Enforcing ZDR requires Pro or Enterprise.

experimental_evaluate or experimental_decide?

Use experimental_decide; the old name is a deprecated alias.

Can I use the TypeSafe SDK with Vercel?

Yes. Set the base URL to https://ai-gateway.vercel.sh/typesafe and use a Gateway key.