Watch AI
pick its sources
Two tiers, both live. The Live Web tier queries engines that actually browse the internet at answer time (Perplexity, Gemini + web, GPT + web). The Pretraining Recall tier asks raw LLMs whether they even know your brand from training data. We only show engines that returned real data.
Live queries, no cache · fuzzy brand match (ignores "Inc", "LLC", &, punctuation) · rate-limited to 5 probes/hour per site. Engines that error or aren't configured are hidden — we only report what actually returned data.
How it works
01
You type a prompt
The kind of thing a real customer would ask an AI.
02
Live Web tier browses
Perplexity Sonar, Gemini + web, and GPT + web each run a real search before answering.
03
Recall tier + fuzzy match
Raw LLMs answer from training data. Then we fuzzy-match against your brand and only report engines that actually returned data.
Live Web tier uses engines that genuinely query the internet at answer time — this is the closest honest proxy for what ChatGPT.com, Perplexity.com, and Google AI Overviews will show a real user. The Recall tier tells you whether the raw underlying model already knows your brand from pretraining.