Gemini Now Sends More Traffic Than Perplexity

A channel you're probably not tracking yet

If you don't have a dedicated segment in Google Analytics 4 for traffic coming from AI engines, you're not alone. Most companies, especially in markets like Romania where we operate, aren't tracking this metric yet. But the numbers from the past few months suggest it's time to start.

Referral traffic from AI platforms was, until recently, a statistical footnote. A percentage so small it didn't justify its own line in monthly reports. That changed visibly at the end of 2025, and by January 2026, the figures are large enough to warrant serious attention.

At difrnt., we've been monitoring AI traffic for all our clients since November. Not because it was significant then, but because we wanted baseline data. What we're seeing now confirms the trends showing up in international studies.

Gemini doubled its referral traffic in two months

A report published by Search Engine Journal, based on SE Ranking data, analyzed referral traffic from AI platforms across a sample of over 101,000 websites with Google Analytics, spanning 250 markets. The findings are notable.

Google Gemini doubled its referral traffic between November 2025 and January 2026. Growth was 51% in December, followed by another 42% in January. Combined, that's approximately 115% growth in just two months.

For context: in the previous 8 months, Gemini had been growing at roughly 4% per month. Then it jumped to 47% per month. A pace 12 times faster than before.

What triggered the acceleration? Three rapid product launches: Gemini 3 Pro (November 18), Gemini 3 Deep Think (December 4), and Gemini 3 Flash (December 17). Flash became the default model in the Gemini app and in AI Mode for Google Search, which explains the surge in users landing on websites through the platform.

From 3x behind to 29% ahead

Perhaps the most striking reversal in the data: in August 2025, Perplexity was sending approximately 3 times more traffic than Gemini. Five months later, in January 2026, the situation had completely flipped.

Gemini now sends 29% more traffic than Perplexity globally. In the United States, the gap is even wider: 41%.

This doesn't mean Perplexity collapsed. It means Gemini grew much faster, pulled by its native integration with the Google ecosystem. When you already have billions of Search users and you ship a series of improved models, the jump is somewhat predictable. But the speed at which it happened caught most observers off guard.

From a practical standpoint, if part of your audience is already using Gemini through Google's AI Mode in Search, that traffic is showing up in your analytics but likely labeled differently from classic organic. Without proper segmentation, you might be confusing it with direct or generic referral traffic.

ChatGPT still dominates, but the lead is shrinking

Let's not lose sight of the bigger picture: ChatGPT still generates approximately 80% of all AI referral traffic. It's the clear leader and will likely remain so for a while.

But the trend matters as much as the absolute number. In October 2025, ChatGPT sent 22 times more traffic than Gemini. By January 2026, that ratio had dropped to 8x. ChatGPT peaked in October, then saw successive declines of 8% in November and 18% in December, with a partial recovery in January.

What does this tell us? That the AI traffic market is diversifying. If your visibility strategy is built exclusively around a single channel, whether that's Google Search or ChatGPT, you risk missing the growth happening elsewhere.

The situation resembles what happened in social media a decade ago: Facebook dominated, but brands that invested early in Instagram and later TikTok gained a competitive edge that was hard to claw back. In AI, the dynamic is similar. ChatGPT is the Facebook of the moment, but Gemini is growing at TikTok-circa-2018 speed.

0.24% of global traffic: small but accelerating

Globally, AI platforms generate 0.24% of total internet traffic (January 2026), up from 0.15% in 2025. Sounds insignificant? Maybe, if you only look at the percentage.

But consider the context: Google generates 94% of organic traffic. The remaining 6% is split among everything that isn't Google. Within that 6%, AI traffic is gaining ground faster than any other source.

For certain niches, the percentage is already much higher. Tech, health, education, and personal finance sites report 1-3% of traffic from AI sources. Within a year, those numbers could look very different across more industries.

What we see with our clients in Romania confirms the trend, even at a smaller scale. For one ecommerce tech client, AI source traffic grew from practically zero in September to 0.8% of total sessions in February. That's not a report-changing percentage, but it grows every month. And it comes with a bonus: conversion rates from AI traffic run 15-20% higher than from classic organic, probably because users arrive with a clearer intent already shaped by the AI conversation.

What to actually do with this information

We recently wrote about how to optimize your content to appear in AI answers. Today's data confirms why that effort is worthwhile: the traffic is coming, and it's growing fast. But optimizing for citations is only half the equation. The other half is measurement.

Here are three steps you can take now:

Set up dedicated segments in Google Analytics 4. Add referral source filters for chatgpt.com, gemini.google.com, and perplexity.ai. If you also use Copilot, add bing.com/chat. Without separate data, you can't evaluate impact or optimize for it.

Check how your site looks from the AI's perspective. Ask Gemini, ChatGPT, and Perplexity questions relevant to your business. See if you appear in the answers, who shows up instead, and what type of content gets cited. You can do this manually or use dedicated tools like Otterly.AI.

Don't treat AI as a channel separate from SEO. Agentic AI engines use quality signals similar to those in SEO: structured content, schema markup, authority through earned media. Optimizing for AI doesn't replace traditional SEO; it builds on the same foundation. An integrated strategy is more effective than two parallel ones.

AI referral traffic is no longer a statistical curiosity. It's a fast-growing channel that's diversifying as more platforms become relevant. The question isn't whether it will matter for your business, but whether you'll be ready to capture it when it matters even more.