ChatGPT traffic drop data shows OpenAI’s chatbot had 5.844 billion web visits in November 2025 (down 5.21% month over month), yet it still ranks as the No. 1 site in the “AI Chatbots and Tools” category and remains a top-five website globally, according to Similarweb.
What happened—and why it matters
ChatGPT’s web traffic dipped in November 2025 after months of heavy usage, falling below the 6-billion-visits mark for that month.
Even with the decline, Similarweb’s category rankings still place chatgpt.com at No. 1 among AI chatbot and tool websites, highlighting that the service continues to lead the standalone chatbot market.
The shift matters because the AI chatbot race is increasingly measured across multiple channels—web, mobile apps, and embedded assistants inside search engines and productivity suites—rather than only by one website’s visits.
The latest traffic snapshot (November 2025)
Similarweb’s November 2025 global top-sites list puts chatgpt.com at No. 5 worldwide with 5.844B visits, down 5.21% month over month.
In the same dataset, gemini.google.com appears at No. 26 worldwide with 1.351B visits, up 14.36% month over month, showing faster growth from a smaller base.
This combination—ChatGPT still far larger in total visits, while Gemini grows faster—captures the core competitive dynamic behind the current “lead despite decline” storyline.
Key AI assistant sites (Similarweb, Nov 2025)
| Site | Global rank (top 100) | Monthly visits (Nov 2025) | MoM change |
| chatgpt.com | #5 | 5.844B | -5.21% |
| gemini.google.com | #26 | 1.351B | +14.36% |
| deepseek.com | #91 | 345.7M | -2.70% |
| chat.deepseek.com | #94 | 330.9M | -2.95% |
“Traffic drop” is not new for ChatGPT
Earlier Similarweb analysis described a prior period when worldwide desktop and mobile web traffic to ChatGPT fell 9.7% from May to June, alongside a 5.7% drop in worldwide unique visitors and an 8.5% decline in time spent.
That earlier dip helped establish a pattern that web visits can soften even while the product remains culturally and commercially significant.
In other words, a monthly decline does not automatically mean demand disappeared; it can also signal normalization after surges and shifting usage to other surfaces.
Timeline of notable dips (Similarweb)
| Period referenced | What changed | Metric |
| May → June (reported by Similarweb) | Web traffic down | -9.7% worldwide |
| May → June (reported by Similarweb) | Unique visitors down | -5.7% worldwide |
| May → June (reported by Similarweb) | Time spent down | -8.5% (decline noted by Similarweb) |
| November 2025 | Web visits down | -5.21% MoM (5.844B visits) |
Why ChatGPT can lead even with fewer visits
Similarweb’s top-sites commentary attributes November’s ChatGPT dip partly to post-launch “normalization” after a spike tied to a major model release earlier in the year.
The same Similarweb analysis also points to intensifying competition from Google’s Gemini and “AI features embedded in search engines,” which can fragment where people go for AI-assisted answers.
At the same time, ChatGPT’s category leadership suggests many users still default to ChatGPT for general-purpose chatbot tasks—especially when they want a dedicated conversational interface rather than an AI layer inside another product.
Engagement signals: strong sessions, not just clicks
Similarweb’s chatgpt.com analytics page reports an average visit duration of 00:06:32, about 3.90 pages per visit, and a bounce rate of 31.68% (as displayed for November 2025).
Those engagement metrics indicate that even if total visits fluctuate month to month, many sessions remain relatively deep compared with quick “one answer and leave” behavior.
For publishers and marketers, that mix—massive reach plus meaningful engagement—helps explain why ChatGPT remains central to SEO and audience strategy conversations even during softer traffic months.
Mobile is reshaping the scoreboard
App-focused research increasingly shows that AI assistant growth is not limited to the web, and that a “web traffic drop” can occur alongside strong mobile adoption.
Start.io’s 2025 app ranking analysis says ChatGPT was the most-downloaded app of 2025 across the Apple App Store and Google Play, with more than 80 million installs and year-over-year growth of more than 100% (as stated in its report).
Separately, Sensor Tower’s “State of AI Apps Report 2025” states that ChatGPT became the fastest app to reach 1 billion global downloads in July 2025.
The competitive map: growth is coming from ecosystems
Similarweb’s November 2025 global list shows Google.com at 82.3B visits and also includes gemini.google.com in the top 100, illustrating how Google can drive AI usage through both search and a dedicated assistant destination.
Similarweb’s notes on Gemini cite tailwinds such as distribution through Google Workspace and Chrome, which can funnel users into Gemini for generation and analysis tasks.
This is a key reason the “AI chatbot market” is now shaped as much by bundling and defaults as by model quality alone, because ecosystems can create habitual usage without users explicitly choosing a standalone chatbot site every time.
What OpenAI emphasizes about usage
OpenAI’s “How people are using ChatGPT” page describes the work as its largest consumer usage study to date and says demographic gaps are shrinking while economic value is being created through personal and professional use.
That framing aligns with the broader market reality suggested by web and app metrics: AI assistants are becoming routine utilities rather than novelty tools.
As usage becomes more “everyday,” volatility in month-to-month web visits can increase because people shift between apps, integrations, and multiple assistants depending on the task.
Final thoughts
A single-month ChatGPT traffic drop does not appear to threaten ChatGPT’s position as the most-visited standalone AI chatbot site, based on Similarweb’s category rank and global placement.
The bigger story is that competition is widening through faster-growing challengers and AI features embedded into existing high-traffic platforms, which may keep standalone chatbot web traffic more volatile.
Going forward, the clearest read on “who is winning” will likely come from combined indicators—web visits, app installs, and ecosystem distribution—rather than web traffic alone.






