Skip to main content

Evaris Documentation

Evaris is a developer platform for running and observing LLM and agent evaluations end-to-end. Use Evaris to:
  • Build evaluation suites with reusable solver and scorer configurations.
  • Run evaluations over your datasets.
  • Inspect traces, events, and sample-level outputs.
  • Compare quality metrics over time.

Product surface

  • Platform app: configure projects, datasets, solvers, scorers, and evaluation suites.
  • Runtime API: submit and monitor runs programmatically.

Where to start

  1. Read Quickstart.
  2. Create your first run with Create your first run.
  3. Explore the generated endpoint reference under the API Reference tab.