Stupid Meter is the world's first AI intelligence degradation detection system. We monitor OpenAI GPT models (including GPT-5, O3, O3-Mini), Anthropic Claude (Opus 4, Sonnet 4), xAI Grok 4, and Google Gemini 2.5 series in real-time.

Our system uses a sophisticated 7-axis scoring methodologycombined with mathematical drift detection algorithms to identify when AI companies reduce model capability to save costs. We track correctness, specification compliance, code quality, efficiency, stability, refusal rates, and recovery ability.

Monitor your favorite AI models including OpenAI GPT-5, Claude Opus 4, Grok 4, and Gemini 2.5 Pro. Get instant alerts when performance drops and track historical trends with our continuous benchmarking system.

Use our "Test Your Keys" feature to run the same benchmarks against your own OpenAI, Anthropic, xAI, or Google API keys for complete transparency. Verify our results independently and see real-time performance metrics.