Gemini hits 1 billion users: how people are actually using Google's AI chatbot

Google's Gemini app has crossed 1 billion users, the company announced, marking a significant milestone for its flagship consumer AI assistant just a few years after its public debut. Alongside the headline number, Google also shared new details about how people are actually using the chatbot day to day — details that paint a picture of an assistant increasingly used by voice rather than text, and one being asked to generate images at enormous scale.
Perhaps the most striking figure in the data Google released is that 63% of Gemini users are talking directly to the assistant using its voice feature, rather than typing out a message the way earlier generations of chatbots required. That's a meaningful shift in how people interact with AI: rather than treating Gemini as a text-based search box or messaging app, a clear majority of its user base now treats it more like a conversational assistant they can simply talk to.
The preference for voice interaction has implications well beyond user comfort. Voice-first usage tends to happen in contexts where typing isn't practical or convenient — while driving, cooking, walking, or multitasking with hands otherwise occupied — suggesting Gemini has moved beyond being a tool people open deliberately to look something up, and closer to becoming a constant companion woven into moments throughout the day.
Alongside the voice statistic, Google disclosed that Gemini now generates more than 150 million images every day. That figure reflects how thoroughly AI image generation has moved from a novelty into an everyday tool, used for everything from quick social media graphics and meme creation to draft visuals for professional design work, marketing material and personal projects that would once have required separate software or design skills entirely.
Reaching 1 billion users represents a scale that very few consumer products in history have achieved, and doing so for a conversational AI assistant marks a new kind of milestone for the industry. It places Gemini among a small handful of Google products — alongside services like Search, Gmail and Android — that have managed to become genuinely global in reach, used by a meaningful fraction of everyone on the internet.
Much of that growth has been driven by Gemini's integration across Google's existing ecosystem of products, giving the assistant a distribution advantage that a standalone app built from scratch wouldn't have. Rather than requiring users to seek out and download a new app in isolation, Gemini has been woven into surfaces many people already use regularly, lowering the barrier to that first interaction that eventually turns into habitual use.
The sheer volume of daily image generation also raises practical questions about the infrastructure required to support it. Generating 150 million images a day, on top of the text and voice interactions from a billion-plus users, demands enormous computing capacity — a demand that sits at the center of the broader, industry-wide race among AI companies to build out data centers and secure the chips needed to keep pace with usage growth.
For everyday users, the growth in voice and image capabilities suggests Gemini is being asked to do more varied work than answering typed questions. A conversational assistant capable of generating images on demand functions less like a search engine and more like a general-purpose creative and informational tool — one that increasingly gets folded into tasks that used to require several different specialized apps.
The milestone also puts Gemini's trajectory under closer scrutiny from industry watchers, who will be looking at whether the pace of growth can be sustained, how usage patterns like the voice-versus-text split continue to evolve, and how the underlying costs of serving this many users at this volume of activity affect the broader AI business model as the technology continues to mature.
For now, the 1 billion user figure stands as a clear signal that conversational AI assistants have moved well past early adopters and into genuinely mainstream daily use. How people are actually using that access — talking more than typing, and generating images by the tens of millions each day — may end up being just as significant a data point as the headline number itself.
Read next

Made by Google 2026: what to expect from today's Pixel 11 launch event
Google is set to unveil a wave of new Pixel devices at its annual Made by Google event on August 12th. Pre-event leaks have offered an unusually detailed preview of what's expected — from a broader Pixel 11 phone lineup to new wearables and possibly a whole new accessory category for the company.

How new surveillance tech links your phone to your license plate
Researchers report that roadside cameras photographing license plates are increasingly paired with equipment that also captures wireless signals from nearby phones. That combination can turn a simple 'this car passed here' record into a much richer profile of who was actually inside.

What are device-bound session credentials, and how do they stop the newest form of account takeover?
Google's Chrome browser has begun rolling out device-bound session credentials, described as among the strongest protections yet against a rapidly growing category of account takeover. The feature targets a technique that steals the session tokens keeping a user logged in, rather than stealing their password.

What is the Riemann hypothesis, and how close is AI to solving it?
On the Riemann hypothesis, one of mathematics' most famous unsolved problems for more than 150 years, an unreleased Anthropic AI model reportedly made significant progress. The model didn't prove the hypothesis, but went further than many experts expected an AI system to get.

What are small language models, and why are they starting to rival giant AI systems?
Liquid AI's new LFM2.5 model, with just 2.6 billion parameters, is competitive with models four times its size. It's the latest example of a fast-growing small language model movement challenging the assumption that bigger is always better in AI.