Before you start

This is a public demo — a live LLM connected to the web. A few things to know:

Capability Ladder — rung 1

Plain LLM vs LLM + web search

Toggle web search on and off mid-conversation. Watch how the same model answers differently when it can look things up — and read why that matters.

RAG soon
Agentic search soon

The model answers from training knowledge only — ask it something recent and watch it hedge or go stale.

Ask anything — try one of these to see the difference:

Enter to send  ·  Shift+Enter for newline  ·  nothing is stored — refresh wipes the conversation