Jev API Examples (Optio)

Copy-paste examples for calling the Jev decision model through Optio, an independently operated access gateway. Jev answers typed questions — choice, score, and yes/no (noul) — with calibrated probabilities. It does not write prose; that is the point.

Optio is an independent gateway. It is not affiliated with, endorsed by, or an official product of TypeSafe. Jev is a trademark of its respective owner.

中文说明

What's here

Path	What it shows
examples/curl.sh	One yes/no decision with curl — the smallest useful request
examples/python.py	Support triage: choice + score + yes/no in one request, with retry and confidence thresholds
examples/javascript.mjs	Same triage shape in plain Node (no dependencies)
skills/jev/SKILL.md	Installable agent skill — npx skills add sunflower225/jev-api-examples (or copy into your agent's skills directory)

Quick start (60 seconds)

# 1. Get a key: register at https://jevai.website — 5 free credits, no card needed
export JEV_API_KEY="jev_live_your_key"

# 2. Make your first decision
curl https://jevai.website/api/v1/systemone \
  -H "Authorization: Bearer $JEV_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "state": "The payments webhook failed twice overnight. Should I wake up the on-call engineer?",
    "questions": {
      "page": {
        "type": "noul",
        "instructions": "Should someone be paged right now?",
        "criteria": {
          "true": "Revenue-impacting or data loss",
          "false": "Can wait until morning"
        }
      }
    }
  }'

The response carries answers keyed by your question names, usage.credits, and latency_ms.

{
  "request_id": "req_fe17d05411092c4f7f97947a",
  "model": "jev-1.13.0",
  "answers": { "page": { "type": "noul", "noul": 0.91 } },
  "usage": { "input_tokens": 128, "output_tokens": 21, "credits": 1 },
  "latency_ms": 432
}

The mental model

• state — the ticket, form payload, or tool call the decision is about. String, array of strings, or JSON object; up to 100,000 characters.
• questions — 1 to 8 named questions. Answer space is declared up front (choice: 2–20 options; score: 2–10 levels), so output always fits your types.
• Your code branches on answers and on the probability — route, block, or ask a human by threshold. Jev judges; your policy stays in code.

Errors always return {error: {code, message, request_id}}. Retry 429 / 502 / 529 with backoff; never retry 401 / 402 / 422 unchanged.

Cost

One credit covers 1,000 input tokens (minimum 1 credit per request; output tokens are free). A typical short decision costs 1 credit — $0.0001 on the $10 Starter pack. Credits never expire. Failed requests are not charged. Full pricing →

Where to try it without a key

The free playground runs the real model in the browser: 30 runs per day anonymously, 50 signed in. The JSON you see there is the JSON your server sends.

Links

• Docs: https://jevai.website/docs
• Playground: https://jevai.website/playground
• Pricing: https://jevai.website/pricing
• Use cases: https://jevai.website/use-cases
• Support: support@jevai.website