In discussions surrounding AI chatbots—also known as large language models (LLMs)—it often feels like ChatGPT dominates the conversation. That perception isn’t just anecdotal; it’s backed by data as well. According to new tracking data from Statcounter, which monitors market share across various digital platforms, ChatGPT holds a commanding lead over its competitors in AI chatbot usage, accounting for more than 80 percent of all sessions tracked globally—and that figure continues to grow. These findings mirror the results of many independent tests and reviews, reinforcing ChatGPT’s status as the clear front-runner in the current AI race.
Statcounter only began including AI chatbot usage in its metrics this past March, but the rankings have remained largely consistent since then. ChatGPT has maintained its dominance, while other contenders—such as Microsoft’s Copilot, Google Gemini, Claude AI, Perplexity, and Deepseek—trail by significant margins. One notable shift occurred with Perplexity AI, which initially captured an impressive 16 percent of U.S. sessions in March before its share fell to just 6 percent by July. In contrast, Microsoft Copilot has made notable gains, increasing from less than 1 percent to nearly 10 percent over the same period.
In the U.S., Statcounter’s July data shows ChatGPT capturing 80.22 percent of all chatbot sessions, followed by Copilot at 9.51 percent and Perplexity at 5.61 percent. Google Gemini accounted for 2.67 percent, Claude AI held 1.56 percent, and Deepseek rounded out the list with 0.43 percent. On a global scale, ChatGPT’s lead grows even further, climbing to 82.69 percent. Perplexity maintains 8.06 percent of sessions worldwide, while Copilot’s share stands at 4.56 percent. Google Gemini (2.2 percent), Deepseek (1.59 percent), and Claude AI (0.91 percent) make up the rest of the usage statistics.
However, some questions remain about how Statcounter is calculating this data. The company claims its analysis is based on over 5 billion page views across 1.5 million websites each month, tracked via cookies that gather device information, OS usage, and referral data. While that provides a strong foundation for measuring browser and search engine traffic, it’s less clear how it applies to LLMs, especially since most AI chatbots don’t refer users to external pages in the traditional sense. That distinction is also raising concerns among content creators, as generative AI tools increasingly act as self-contained sources of information rather than conduits to the open web.
Statcounter has yet to respond to questions about its methodology for chatbot usage tracking.
