Skip to content
Rasmus Wulff Jensen edited this page Sep 22, 2026 · 3 revisions

JevDotNet wiki

JevDotNet turns answers from the TypeSafe AI Jev API into a C# object you define. You can ask several typed questions about the same text in one call.

New here? Start with Getting started. It takes you from a new .NET project to a working EvaluateAsync<T>() call.

Implement in this order

  1. Install and make your first call. Add the NuGet package, supply an API key, and run a complete example. JevDotNet targets .NET 8 or later.
  2. Build the return object. Decide what your code needs to know, then add one attributed, writable property per question.
    • Choice selects one named option. Use it for categories or routing.
    • Score places an answer on an ordered scale. Use it for severity or degree.
    • Noul gives the probability of yes. Use it for a proposition or Boolean decision.
    • Compare property types when choosing simple values or detailed answer records.
  3. Evaluate text and use the result. Read Result and token counts, inspect confidence or probabilities when needed, and configure the client for your application.
  4. Troubleshoot. Use the property name and linked guide in an exception to fix return-object, configuration, or response-shape errors.

The repository README is a short reference; this wiki is the implementation walkthrough.

Clone this wiki locally