typesafe-aiCertifiedReleased 09/15/2026

Jev

TypeSafe AI System One decision model. Accepts shared state and typed questions, returning choices, scores, and boolean probabilities. Limits: 64,000 tokens total per request; 32,000 tokens for the state plus the longest question. No streaming or output-token limit.

View API reference ↓
Input / output priceFrom$0.042 / $0per 1M
24h uptime100.00%
const res = await fetch("https://api.inferencebureau.com/v1/evaluate", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.IB_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "typesafe-ai/jev",
    "state": { "review": "Arrived early, works great." },
    "questions": {
      "sentiment": {
        "type": "choice",
        "instructions": "Overall sentiment of the review?",
        "criteria": { "positive": "Favorable", "negative": "Unfavorable" }
      }
    }
  }),
});

const { answers } = await res.json();
console.log(answers.sentiment.choice, answers.sentiment.confidence);

Providers

Dispatch routes requests across these providers and fails over between them without changing the model. Restrict or order providers with team Directives. Prices are gateway list prices before the platform fee.

ProviderContextMax outputLatencyThroughputInputOutputCacheCapabilitiesZDRNo trainingRegionsStatus
digitalocean32K0——$0.042$0Read —Write ——NoYesGlobalCertified
typesafe-ai32K0——$0.042$0Read —Write ——YesYesUSCertified

Uptime

24 hours

Request success rate per provider.

digitalocean100.00%
typesafe-ai100.00%

Getting started

Create a key on the API Keys page and set it as IB_API_KEY. See the quickstart.

const res = await fetch("https://api.inferencebureau.com/v1/evaluate", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.IB_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "typesafe-ai/jev",
    "state": { "review": "Arrived early, works great." },
    "questions": {
      "sentiment": {
        "type": "choice",
        "instructions": "Overall sentiment of the review?",
        "criteria": { "positive": "Favorable", "negative": "Unfavorable" }
      }
    }
  }),
});

const { answers } = await res.json();
console.log(answers.sentiment.choice, answers.sentiment.confidence);

Jev answers POST /v1/evaluate, not chat completions. It is billed on input tokens only.

More models by typesafe-ai

ModelContextInputOutputCacheRelease
typesafe-ai/jev-1.14.032K$0.042$0Read —Write —10/15/2026
typesafe-ai/jev-1.13.032K$0.042$0Read —Write —09/15/2026