How chat works

Every chat message is routed to the right reasoning engine based on what you asked — fast for casual exchanges, deep reasoning for hard problems, live-grounding (real web + X search) for anything time-sensitive. On any plan you can pick an engine per bucket in Settings → Models.