Apple put a number on Siri’s daily workload for the first time this week. Lilian Rincon, Apple’s vice president of product management and marketing for artificial intelligence, told the September 9 Apple Event keynote that “our users make over 2.5 billion requests every single day.” The figure matters to anyone tracking how much query volume now flows through voice and AI assistants rather than a traditional search box.

Rincon’s exact line, delivered around the 9

mark of the keynote, was: “In fact, our users make over 2.5 billion requests every single day. And our new version of Siri goes further than ever before.” She followed with a claim that Siri is “already one of the most widely used assistants in the world,” a description Apple did not back with a ranking or a named comparison set.

That absence is the point for anyone sizing this stat. Apple did not say how 2.5 billion requests compares to Google, Alexa, or any other assistant, and the company has not published a prior daily figure for Siri to measure this one against. The number lacks public context: it is large on its face, but nothing in Apple’s presentation lets an outside observer judge whether it represents growth, stagnation, or simply the first time Apple chose to disclose it.

Search Engine Roundtable, which first reported Rincon’s remarks, published the keynote slide showing the figure alongside the quote and noted that Apple’s own framing invites an obvious comparison Apple did not make itself. The publication’s own writeup flagged Google as the presumed larger player without offering a competing figure, underscoring that no verified head-to-head number currently exists in the public record.

For search and GEO practitioners, the disclosure is still useful as a scale marker rather than a competitive scoreboard. Siri sits inside Apple’s ecosystem of iPhone, HomePod, CarPlay, and Mac devices, and 2.5 billion daily requests confirms that answer-style, voice-driven queries now represent a query volume large enough that Apple is willing to state it publicly for the first time. That is a signal about where discovery behavior is heading even without a benchmark to compare it against.

The disclosure also arrives as Apple has been clarifying how its AI crawlers interact with the rest of the web, including recent confirmation around Applebot-Extended, the crawler token Apple uses to gather training data separately from standard indexing. Whether Siri’s answer layer draws on the same content signals that determine classic Search rankings, or draws on a separate pipeline entirely, remains a question Apple’s keynote did not address.

Teams that optimize for answer engines should treat this figure as confirmation that Apple’s assistant now operates at a scale worth monitoring, not as evidence of a specific market share shift. Anyone running GEO, generative engine optimization aimed at how LLM-driven assistants surface answers, or AEO, answer engine optimization for direct-answer surfaces, should watch for Apple to publish a follow-up figure or a year-over-year comparison, since that second data point would be the one that actually establishes a trend.

Search Engine Roundtable reported Lilian Rincon’s remarks on September 10, 2026, following her appearance at Apple’s September 9 keynote.