JevModel

What is JevModel?

JevModel is an independent service that puts a structured decision model behind a web console and an API. You give it a state, such as a support message or a comment, together with focused questions, and it returns typed answers that your application can check and act on. It is aimed at developers and operations teams that need to route, label, or escalate content and messages in a repeatable way. The site states that it is an independent service and is not affiliated with TypeSafe AI, the provider of the underlying model.

How to use JevModel?

1. Create an account and verify your email, which unlocks five free runs every UTC day.
2. Open the playground and write a question, choosing one of three answer types: Choice, Score, or Noul.
3. Test the question on your own examples and review the probability attached to each answer.
4. Save the decision in a flow, or start from a template such as support inbox triage, comment moderation, or lead qualification.
5. Create an API key in the console and call the same decision from your own application.

Key features of JevModel

• Three answer types: Choice selects one of your options, Score assigns a level from your scale, and Noul makes a yes or no judgement against a criterion
• Decision flows drawn on a canvas, where each answer leads to the next question or to an outcome
• A single request that runs a whole flow, billing only the questions on the selected path
• Batch processing that accepts CSV, TXT, or JSON Lines files of up to 1,000 rows and exports row-level answers

Use cases for JevModel

Content moderation

Hold back insults, harassment, and threats before they are published.

Support triage and routing

Work out what a customer wants before anyone opens the ticket, then send the message to the right team or queue.

Pricing for JevModel

• Free: five successful runs every UTC day after email verification, subject to site capacity
• Starter: $9.90 per month for 30 million tokens
• Builder: $19.90 per month for 80 million tokens
• Pro: $39.90 per month for 200 million tokens, and Scale: $79.90 per month for 500 million tokens

Frequently asked questions about JevModel

Is a probability a guarantee?

No. The site advises testing examples from your own workflow and reviewing borderline results, because a model probability is not a guarantee that an answer is correct.

What are the request limits?

A normal request supports up to eight questions. Flows support up to eight question nodes and 24 total nodes, with no cycles.