Apigator by MixerBox

TypeSafe Jev · on Apigator

Decisions, not chat.

Jev is a decision model. Give it any text or JSON plus typed questions; get back structured answers with calibrated probabilities in ~100 ms. No prompts to babysit, no output to parse.

Sign up free — get 1,000,000 Jev tokens on us

Every new Apigator account gets it automatically. Already have an account? Sign in through this page to claim it once.

Share Jev

Why Jev

🧩

Typed answers

Every answer comes back as JSON keyed by the questions you named — a choice, a score, or a yes/no probability. Your code reads it directly.

📊

Calibrated probabilities

Each answer carries a probability and a confidence. Act automatically when it is sure, send the rest to a human or a bigger model.

~100 ms

Fast enough to sit inline in a request path: routing, moderation, filtering, ranking.

🗂️

Many questions, one call

Ask department, urgency and sentiment about the same text in a single request — one round trip, one bill.

💸

Very cheap

$0.042 per 1M input tokens, output free. Classify millions of items for the price of a coffee.

🔑

Same Apigator key

No new account or contract. Your existing Apigator key and balance work for Jev and every other model.

Three question types

"type": "choice"

Choice

Pick one option from a set you define (up to 255). Returns the pick plus a probability for every option.

"type": "score"

Score

Rate against ordered levels you describe (2–10). Returns a probability-weighted score that can land between levels.

"type": "noul"

Noul (yes/no)

Returns the probability that a yes/no statement is true, from 0 to 1.

Jev vs. an LLM chat model

Chat models write. Jev decides. Use each for what it is good at.

LLM chat (GPT, Claude, …) Jev
Output Free-form text Typed JSON answers, one per question
Parsing JSON mode, regex, retries when it drifts None — the shape is guaranteed
How sure is it? No reliable signal Calibrated probability + confidence on every answer
Latency Seconds ~100 ms
Price Pay for input and output tokens $0.042 / 1M input tokens, output free
Best for Writing, reasoning, conversation Classify, route, score, filter, guardrails

💡 Common pattern: Jev screens and routes every request, and only the ones that need writing go to an LLM.

Example: triage a support ticket

One request asks three questions about the same message.

Request
curl https://api.apigator.ai/typesafe/v1/systemone \
  -H "Authorization: Bearer $APIGATOR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "jev-latest",
    "state": "Help! My payouts have been failing for 3 days.",
    "questions": {
      "department": {
        "type": "choice",
        "instructions": "Which team should handle this?",
        "criteria": {
          "billing": "Payments, invoicing, refunds",
          "technical": "Bugs, outages, integrations",
          "sales": "Pricing, upgrades, new accounts"
        }
      },
      "is_urgent": {
        "type": "noul",
        "instructions": "Does this convey urgency?"
      },
      "frustration": {
        "type": "score",
        "instructions": "How frustrated is the customer?",
        "criteria": ["Calm", "Frustrated", "Very angry"]
      }
    }
  }'
Response
{
  "model": "jev-1.13.0",
  "answers": {
    "department": {
      "type": "choice",
      "choice": "billing",
      "probabilities": { "billing": 0.88, "technical": 0.12, "sales": 0.0 },
      "confidence": 0.82
    },
    "is_urgent": { "type": "noul", "noul": 0.95 },
    "frustration": {
      "type": "score",
      "score": 1.04,
      "legend": { "0": "Calm", "1": "Frustrated", "2": "Very angry" },
      "probabilities": { "0": 0.0, "1": 0.96, "2": 0.04 },
      "confidence": 0.94
    }
  },
  "usage": { "input_tokens": 402, "output_tokens": 73 }
}
Then just use it in code
const { answers } = await res.json();

if (answers.is_urgent.noul > 0.8) pageOnCall();
if (answers.department.confidence > 0.7) {
  assignTo(answers.department.choice);   // "billing"
} else {
  sendToHumanTriage();
}

What people build with it

Support ticket routingContent moderation & LLM guardrailsRAG passage filteringLead / resume scoringIntent routing for agentsCitation & duplicate checks

Pricing

$0.042

per 1M input tokens

Output tokens: free

A call like the example above uses 300–400 input tokens (measured), so the free 1M tokens cover roughly 2,500–3,000 calls. After that, Jev draws from your normal Apigator balance — pay as you go, no subscription.

Sign up free — get 1,000,000 Jev tokens on us

Get started in 3 steps

  1. 1

    Sign up with the button on this page — 1M Jev tokens are added to your balance automatically.

  2. 2

    Copy your API key from the dashboard.

  3. 3

    POST to https://api.apigator.ai/typesafe/v1/systemone — or use the official SDK with the two environment variables below.

SDK
pip install typesafe-sdk      # or: npm install typesafe
export TYPESAFE_BASE_URL=https://api.apigator.ai/typesafe
export TYPESAFE_API_KEY=sk-your-apigator-key

References

Try Jev free

1M tokens, no card needed.

Claim 1M free tokens
Share Jev