Jev Model API is a specialized tool that turns text into structured, typed decisions for software applications. Developers use it to build logic gates that require intelligence but need to return predictable JSON instead of long paragraphs of text. It solves the common problem of prompt engineering and text parsing by forcing the AI to choose from a predefined vocabulary you set.
The system works as a fast decision layer. You provide a context, which the tool calls the state, and ask specific questions. The API then returns a label, a confidence score, and a reason. This makes it reliable for production workflows where you need to route a ticket or block a comment without worrying about the AI hallucinating a conversational response.
Key features
- Returns structured JSON objects with labels and routes for easy integration
- Uses the jev-1.13 model specifically tuned for classification tasks
- Supports multiple independent questions within a single context state
- Consumes exactly one credit per API call regardless of question count
- Provides a confidence score for every decision to help with auditing
- Includes a web-based playground for testing logic without an account
- Offers a stable API endpoint at /api/v1/decisions for production use
How to use
- Open the Jev Model playground to test your logic for free without signing up.
- Input the state which acts as the shared context for your questions.
- Define your questions and the specific vocabulary of answers you want the AI to use.
- Run the decision to see the JSON output and check the confidence levels.
- Sign in with Google to generate a permanent API key in your account settings.
- Integrate the POST request into your application code using the provided endpoint.
Use cases
- Automating support ticket triage by labeling requests as normal, priority, or escalate.
- Filtering user-generated content by deciding to approve, block, or send for human review.
- Routing chatbot intents to specific departments like billing, sales, or technical support.
- Assessing risk for incoming requests to decide if they should be allowed or rate-limited.
Pricing
Jev Model offers 1,000 free credits daily. One-time credit packs are available starting at $10 for 100,000 credits, $100 for 1,000,000 credits, and $1,000 for high-volume needs. Credits are valid for 365 days from purchase.
FAQ
What is Jev Model API?
It is a typed decision layer that returns structured data instead of conversational text for application logic.
Is Jev Model API free?
Yes, the service provides 1,000 free credits every day without requiring a credit card.
How are credits measured?
Each decision call consumes 1 credit, no matter how many questions are included in that single request.
Is this a chatbot?
No, Jev Model is a classification tool meant for labels, routes, and scores rather than conversation.
What does a response look like?
Every response is a JSON object containing a decision label, a confidence score, a reason, and an optional route.
Can I use it for API resale?
Yes, the system includes credit metering and stable endpoints specifically to support high-volume and resale use cases.
