AMPD queries 9 AI models to measure how brands appear in AI answers, plus a simulated Google AI Overviews view. Every surface is responding normally. Last checked 2026-07-29 10:07 UTC.
Every hour we send a live query to each platform and record whether it answered, and how long it took. Uptime is the share of those checks that succeeded.
Checks that failed because of a fault on our side — a misconfigured probe or an out-of-date model identifier — are excluded. They say nothing about whether the platform was available, and counting them would understate a provider's reliability.
One entry is marked simulated. Google AI Overviews has no public API, so we approximate it by prompting the Gemini model for an Overviews-style answer. It is a useful view, but it is not a separate platform — so we count 9 independently measured models, not 10.