🔗 LLM Integration Lab

Type a query, watch it get embedded and compared against a small document corpus, and see which documents get retrieved and injected into the prompt as grounding context.

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install