| # | Model | Score | Symptoms | OK | Issues |
|---|
Each symptom is grounded in a peer-reviewed paper.
AI Clinic is a diagnostic framework for LLM behavioral health. Instead of measuring capability, it measures mental health — 119 symptoms across 15 dimensions, each grounded in peer-reviewed research.
Each symptom is tested via A/B comparison (control vs experimental prompt) with LLM-as-judge scoring and Wilson 95% confidence intervals.
Open source: Apache 2.0. GitHub
AI Clinic is open-source (Apache 2.0). Cloud features are optional.