Input

Server mode uses the llama.cpp integration. Browser mode runs the Bonsai model locally in your browser and sends only structured findings back to FastAPI for deterministic anonymization.

Results

Results will appear here after processing...