JevModel
VisitDecisions, not text: typed answers from the Jev model via playground and API
JevModel introduction
JevModel is a hosted playground and API for the Jev decision model, which classifies text or JSON into typed answers with probabilities instead of free-form text.
- Website:
- jevmodel.app
JevModel overview
JevModel offers three question types: Choice (pick one of your options), Score (a level on your scale) and Noul (yes or no against a criterion). You can try it in the browser, chain questions into visual decision flows that run with one API request, batch-process up to 1,000 rows from CSV, TXT or JSONL files with CSV export, or call the REST API with a Bearer key. Live runs call typesafe/jev-1.13 through OpenRouter; the site states it is independent of TypeSafe AI. Token packs never expire and failed requests are not charged.
JevModel features
- Choice, Score and Noul question types with probabilities
- Visual decision flows run with one API request
- Batch processing up to 1,000 rows with CSV export
- REST API with Bearer key authentication
- Console for keys and balance
- Token packs never expire; failed requests are free
Questions about JevModel
JevModel pros
- Structured, typed outputs instead of free text
- Free tier with no signup
- Batch and API options
JevModel limitations
- Narrow focus on classification tasks
- Depends on a third-party model via OpenRouter
JevModel use cases
- Moderate user comments
- Filter spam and scam messages
- Triage and route support tickets
Who JevModel is for
- Developers
- Support teams
- Trust and safety teams
JevModel pricing
3 free no-signup runs per day and 100,000 free tokens after email verification. One-time token packs: Icebreaker $4.90 (8M), Starter $9.90 (20M), Builder $19.90 (50M), Pro $49 (180M), Scale $99 (450M). Monthly and yearly plans are also offered.









