ChatGPT’s search-like query volume is about one-fifth of Google Search Volume, according to OpenAI consumer product lead Adam Fry, who said ChatGPT counts a search whenever a question sends it to the web. That comparison describes search requests inside ChatGPT, not the visits ChatGPT sends to websites.
Fry shared the estimate during a Monday evening dinner in New York hosted by OpenAI. Alex Kantrowitz of Big Technology asked how the company counts searches in an open discussion. Big Technology posted the transcript on Tuesday.
Google Search Volume Depends on the Measure
Fry said each question that sends ChatGPT to search the web counts as a search. He described growing use for current information about people, places and things, as well as finance and health questions. Earlier, he said, people mostly used ChatGPT for writing and coding help, which they would not have searched for.

Fry did not identify the Google figure behind the comparison, the time period it covers or the regions included. The source also does not establish whether ChatGPT searches replace searches on Google. The estimate should therefore be read as OpenAI’s account of its own web-search activity, compared with an unspecified Google measure.
The distinction matters for companies comparing search behavior with website referrals. Search Engine Journal’s report on the comparison says outside traffic data measures a different outcome: visitors arriving at sites.
Referral Data Tells a Different Story
Ahrefs estimated in February that ChatGPT search volume was around 12 percent of Google’s, using a different definition. Its estimate started with 2.5 billion daily ChatGPT prompts in July 2025, based on a study by Harvard University and OpenAI. Ahrefs estimated that approximately 65 percent were queries users would previously have taken to Google.
Ahrefs compared that estimate with more than 5 trillion searches per year that Google announced in March 2025. The article notes that the 65 percent share exceeded the portion researchers had classified as search. The approaches differ: one sorts prompts by request type, while Fry’s description counts requests that trigger a web search.
Ahrefs also tracks referrals across sites using its free Web Analytics tool. In October, Google accounted for 24.74 percent of tracked traffic and ChatGPT for 0.36 percent across 118,671 sites, according to the tracker’s Nov. 3 update. Those percentages describe tracked website traffic, not the total number of searches on either service.
For related context on how AI discovery can affect site visits, see this report on AI search and website traffic. A separate account of Google’s AI search links covers changes to how its results direct people to websites.
Related findings from the State of SEO 2026 report examine how teams assess content and search results.
Google Reports Growth as a Rate
Alphabet’s third-quarter release gave Google search query growth as a rate rather than a count. CEO Sundar Pichai said AI investments were driving overall search query growth and called the year-over-year rate the highest in five years. That statement does not provide a direct volume figure for comparison with Fry’s estimate.
Google’s reported growth rate, OpenAI’s search-like query comparison and Ahrefs’ referral shares each describe different measures. A company planning channel reports can keep those measures distinct by stating whether a figure counts searches, prompts or visits sent to a website. Ahrefs’ month-by-month referral tracking offers one recurring outside view, while a closer comparison with Google remains difficult until OpenAI clarifies the figure Fry referenced.



