Llama.cpp extension compares local model performance

This essential "quality of life" tool for the local LLM community transforms ephemeral console logs into actionable hardware optimization data. It intercepts SSE streams directly from the browser's fetch API to ensure compatibility across various web-based UIs while using local IndexedDB storage for privacy. The extension features side-by-side model rankings and scatter plots to determine optimal performance ratios, alongside JSONL export and PNG dashboard stitching for deeper analysis.