ChatGPT Sees 22% Traffic Drop as Gemini 3 Rollout Reshapes AI Demand

Web analytics data shows that traffic to ChatGPT has declined by about 22% in the six weeks following the launch of Google’s latest AI model, Gemini 3, suggesting a shift in user behaviour within the rapidly evolving AI landscape. Analysts caution that broader factors like seasonal trends and distribution differences also influence these figures.

Introduction

Recent web traffic data indicates a notable drop in visits to the ChatGPT web platform in the weeks after Google rolled out Gemini 3, its newest generative AI model. According to data shared by analytics sources, ChatGPT’s seven-day average visitors fell from around 203 million to about 158 million—a roughly 22 % decline over approximately six weeks. Analytics Insight+1

While ChatGPT remains a dominant AI service with a large user base, the change highlights how competition and distribution advantages can affect engagement patterns in the fast-moving AI sector.

What the Traffic Drop Means

Analytics Snapshot

• Web visits to ChatGPT have reportedly decreased from ~203M to ~158M in a six-week span following the Gemini 3 launch, according to SimilarWeb-sourced figures shared publicly.
• Over the same period, Gemini’s web traffic has remained more stable, and its audience size is now estimated to be close to 40 % of ChatGPT’s web volume.

These figures focus on web visitation and don’t account for usage within native apps (iOS, Android) or API-driven integrations, where patterns may differ.

Influence of Competition

Google’s deployment of Gemini 3 across its ecosystem—including integration with Search, Android, and Workspace—gives it unique distribution reach that may encourage users to try or adopt the service without directly navigating to a separate website. Analysts note this structural advantage can play a role in shifting traffic patterns.

By contrast, ChatGPT’s web traffic metrics reflect navigation to the chatbot’s site specifically; users interacting through apps or platforms may not be captured in these web visit counts.

Beyond Numbers: What Else Is Driving Change

Seasonal and Broader Trends

Part of the traffic shift coincides with the end-of-year holiday period, which traditionally sees reduced online activity as users engage less with productivity tools and services. This makes interpreting short-term analytics more complex and suggests traffic drops may not solely reflect competitive displacement. Analytics Insight

Market Fragmentation

The generative AI landscape is now highly competitive, with several notable players—such as Anthropic’s Claude, xAI’s Grok, and open-source alternatives—offering diverse capabilities. This fragmentation can distribute user attention across multiple platforms, diluting traffic to any single service.

Usage Outside Web Metrics

It’s also important to consider that web traffic statistics do not capture usage via mobile and desktop apps, embedded experiences (e.g., in productivity suites), or API usage by businesses and developers. These channels represent substantial proportions of total engagement that aren’t reflected in SimilarWeb’s web visitation data.

Implications for OpenAI and the AI Market

Competitive Dynamics

The apparent dip in ChatGPT’s web traffic comes amid escalating competition between major AI providers. Google’s strategy of embedding Gemini deeply into its core products contrasts with OpenAI’s focus on a standalone platform and third-party integrations.

Despite the dip, ChatGPT continues to maintain a large user base; independent estimates suggest weekly active users in the hundreds of millions. DemandSage

User Behaviour and Stickiness

Traffic metrics alone don’t fully capture engagement quality or user satisfaction. Some analysts argue that user preferences may be driven by factors such as model performance, feature ecosystems, or real-world utility. Additionally, adoption patterns can vary widely by region, use case, and demographic.

These dynamics indicate that overall market share and engagement encompass more than web visits alone, and AI platforms will likely continue evolving their products and distribution strategies in response.

Conclusion

The reported 22 % drop in ChatGPT web traffic following the rollout of Gemini 3 points to shifting user behaviour in the generative AI space, influenced by competitive offerings, distribution reach, and seasonal trends. While web visits offer one lens on engagement, a complete picture of usage across apps, APIs, and embedded experiences is essential to understanding broader platform dynamics.

As major AI providers refine their models and ecosystem strategies, competition is poised to intensify, shaping how and where users interact with artificial intelligence moving forward.

Key Highlights

• ChatGPT’s web visits declined approximately 22 % after the Gemini 3 release.
• Gemini’s web audience has remained steadier, approaching about 40 % of ChatGPT’s web volume. LinkedIn
• Seasonal effects and competitive fragmentation complicate interpretation of traffic data.
• Web metrics capture only part of overall engagement; app and API usage remain significant factors.