LLM Eval Harness

Quick test

Try one prompt on several models at once and compare the answers side by side. Nothing is saved.

Test many inputs instead
What this does: sends your prompt to each AI you pick, then a grader AI scores every answer from 1 to 10 and says why. Use it to see which AI handles your task best. Fill in the boxes (or click an example), pick 2–3 AIs, and press Run.
Start from an example:

Write {{input}} where the input should go, or leave it out and just ask a question.

Grading and more options

Another AI scores each answer from 1 to 10 and explains why.

No models are available. Add an API key to the backend settings file (free Gemini and Groq keys work) and restart the backend.

Pick at least one model. ⌘ Enter