5 free API runs per UTC day
Jev APICall the Jev model with one HTTP request
Send a state and up to 8 typed questions to POST https://jevmodel.app/api/v1/systemone with a JevModel API key. You get Choice, Score and Noul answers as JSON that your code can check, from the live typesafe/jev-1.13 model.
Jev API quickstart
01
Create a key
Sign up, verify your email and create an API key in the Console. Keep it on your server, never in browser code.
02
Send a request
POST JSON with a state (text, a JSON object or a JSON array) and 1 to 8 named questions. Do not send a model field; the model is fixed.
03
Read the answers
Answers are keyed by question ID. Choice returns a label and probabilities, Score a number, and Noul a probability from 0 to 1.
curl -X POST https://jevmodel.app/api/v1/systemone \
-H "Authorization: Bearer $JEVMODEL_API_KEY" \
-H "Content-Type: application/json" \
-d '{"state":"The customer was charged twice and wants a refund.","questions":{"queue":{"type":"choice","instructions":"Which team should handle this message?","criteria":{"billing":"Payments, refunds or invoices.","technical":"Bugs, errors or outages.","sales":"Plans, quotes or upgrades."}}}}'{
"code": 0,
"message": "ok",
"data": {
"request_id": "example-id",
"answers": {
"queue": {
"type": "choice",
"choice": "billing",
"probabilities": {
"billing": 0.93,
"technical": 0.04,
"sales": 0.03
}
}
},
"usage": {
"inputTokens": 142
}
}
}Jev API pricing
The first 5 successful runs each UTC day are free after email verification, subject to site capacity. After that, each successful request is billed on the input tokens the model reports; output is free and failed requests are refunded. Token packs never expire, and Playground, Batch and API share one balance.
See plans and token packs →Limits and errors
- ✓State and full request: up to 8,000 characters each.
- ✓1 to 8 questions per request. Choice takes 2 to 12 options; Score takes 2 to 10 ordered levels.
- ✓Errors return {"code":-1,"message":"ERROR_NAME"} with HTTP 400, 401, 403, 413, 429, 502 or 503.
- ✓Saved Decision Flows run with POST /api/v1/flows/{id}/run and need a paid token balance.
JevModel API or the official TypeSafe API?
TypeSafe AI offers its own Jev API at api.typesafe.ai with first-party Python and JavaScript SDKs. The JevModel API is an independent HTTP endpoint for the same model, served through OpenRouter.
| TypeSafe API | JevModel API | |
|---|---|---|
| Endpoint | POST api.typesafe.ai/v1/systemone | POST jevmodel.app/api/v1/systemone |
| Key | TypeSafe console key | JevModel key from the Console |
| SDKs | typesafe-sdk (Python), @typesafe-ai/sdk (JS) | Plain HTTP; curl, Node.js and Python examples on GitHub |
| Free usage | See TypeSafe pricing | 5 free successful runs per UTC day |
| Extras | First-party support and SDK updates | Shared balance with Playground, Batch and saved Decision Flows |
Choose the TypeSafe API if you want first-party SDKs and direct billing. Choose JevModel if you want free daily runs, a no-code Playground and Batch on the same balance. Keep the request in one adapter so you can switch later.
Jev API FAQ
Is there a free Jev API?
Yes. After email verification, the JevModel API includes 5 free successful runs per UTC day, shared with the Playground. Additional runs use your token balance.
How much does the Jev API cost?
Successful requests beyond the free runs deduct the model-reported input tokens from your balance. Output tokens are free. Current plan and pack prices are on the pricing page.
Is there a Jev SDK or GitHub example?
JevModel does not ship its own SDK; any HTTP client works. The jevmodel-examples repository on GitHub has curl, Node.js and Python clients. TypeSafe publishes official SDKs for its own API.
Is the JevModel API affiliated with TypeSafe AI?
No. JevModel is independent and calls typesafe/jev-1.13 through OpenRouter. A JevModel key does not work on api.typesafe.ai, and a TypeSafe key does not work here.
Can I send images or files?
No. The API accepts text or JSON. For images, send captions or OCR text you already have.