There is one clear leader and a competitive chasing pack. Measured by unique monthly visitors to their websites, ChatGPT sits far in front at roughly 489 million, with Gemini next at about 122 million, then DeepSeek, Grok, Perplexity, Copilot and Claude trailing in the tens of millions. On mobile apps the order shifts a little, but ChatGPT still leads by a wide margin at around 557 million monthly active users.
Audience is not the same as referral
Here is the nuance that matters for visibility. The size of an assistant's audience is not the same as how much traffic it sends back to the open web. On measured referral traffic the concentration is even sharper: ChatGPT accounts for roughly 81 percent, Perplexity for about 8 percent, and Copilot for around 5 percent, with the rest sharing what is left. So the assistant most likely to send a real visitor to your site is, by a distance, the same one most people use.
Perplexity is worth a second look despite its smaller audience, because it is built around citing its sources visibly, so being quoted there is both more likely and more useful. The mechanics of who gets named are the same across all of them, and we cover them in how assistants choose which brands to recommend.
Where to spend limited effort
You do not have to court a dozen systems at once. The leaderboard tells you to be unmistakably legible to the handful that people actually use, and clearest of all to ChatGPT, because that is where both the audience and the referrals concentrate. Get the fundamentals right and you tend to become readable to all of them together, since they draw on much of the same open web.
What this means for your business
Do not chase every new model that launches. Anchor your effort to the leaders, confirm the assistants can read and describe you correctly, and treat the citation-forward tools like Perplexity as a useful proving ground for whether your content is quotable. The scale of AI adoption means even second-tier assistants now reach audiences larger than many national markets.