Hi! This chat runs a small LLM in your **browser** (WebLLM). The first reply may take a bit while the model downloads.