Engines

The four engines, one at a time

ChatGPT, Perplexity, Gemini and Google AI Overviews read differently, cite differently and are asked in different places. What each does, and what to do about it.

Updated September 2026 · by Monish

The four engines the report scores

EngineCompanyReads withShows sources as
ChatGPTOpenAIOAI-SearchBot, ChatGPT-User, GPTBotLinks in the answer, attached to the sentences they support, and a sources list.
PerplexityPerplexity AIPerplexityBot, Perplexity-UserNumbered sources, shown with the answer, one per claim.
GeminiGoogleGooglebot, Google-ExtendedGrounding sources, returned with the answer as a set of URLs, each behind a Google redirect.
Google AI OverviewsGoogleGooglebotLinks beside the text, one set per overview, and a sources panel.

Every question in a report goes to all four, with the buyer's country set, and the report says per engine who was named and what was cited. An engine that could not be scored is named in the header with the cause.

Where they differ

ChatGPT and Perplexity search on every asking and assemble the list from what they fetched; Gemini's grounded answer follows Google's ranking; AI Overviews are written from the ranked pages and are not shown for every question. A company can be strong on one and absent on another, which is why the report scores them separately and shows the split.

The crawlers differ too. OpenAI and Perplexity run their own; Google's engines read the Search index, with Google-Extended as the opt-out for Gemini and the ordinary snippet controls for AI Overviews. A robots.txt written in 2023 to block 'AI' often blocks the one that matters and leaves the others.

Read next

By industry · By country · Check your robots.txt

See what each engine says about you

Your website, region and work email. The score appears in about three minutes, and the report is a link you keep.

Check my AI visibility