JevModel

Decision flows let you chain decisions on a canvas — spam check, intent, routing, urgency — where each answer leads to the next question or an outcome. One request runs the whole flow, and only the questions on the path are billed.

Batch processing accepts CSV, TXT or JSON Lines files of up to 1,000 rows, runs a template or your own question over them, and returns row-level answers with CSV exports so you can review results before choosing a production threshold.

One API uses a single request for Choice, Score and Noul answers. You create a key in the Console and send requests to the same account balance used by the playground and batch jobs. Python and JavaScript examples are provided, and the API reference covers request types, probability fields and error handling.

The page lists decision tasks such as content moderation to hold back insults and threats, spam and scam filtering for inboxes and forms, support triage to identify what a customer wants, and ticket routing to the right team.

Other examples include priority and escalation for outages, language detection before choosing a response, personal data screening before storing or forwarding text, prompt injection screening before an AI agent reads web pages or tool output, and sentiment tracking across tickets, reviews and chats. Templates for support inbox triage, comment moderation and lead qualification are available as starting points and can be modified.

JevModel offers 3 live runs without signup, followed by 100,000 free tokens after email verification. The page states that you pay for tokens and keep what you do not use, and that the Console, batch jobs and API all draw from the same account balance.