Google has revealed new numbers showing how quickly its artificial intelligence products are expanding across consumers, developers and enterprise customers.
During Alphabet’s second-quarter 2026 earnings announcement, CEO Sundar Pichai said the Gemini app had reached 950 million monthly active users. Daily active users have also tripled over the past year.
The company reported similarly strong growth in its cloud business, where demand for artificial intelligence infrastructure, enterprise agents and Gemini-powered services helped Google Cloud revenue increase by 82% compared with the same period a year earlier.
Google also confirmed that it has started what it describes as its most ambitious model pre-training run yet: Gemini 4.
The results demonstrate that the competition between Google, OpenAI, Anthropic and other AI companies is no longer limited to model benchmarks. The companies are now competing for users, developers, enterprise customers, data center capacity and the computing power needed to operate increasingly capable AI systems.
Gemini Reaches 950 Million Monthly Users
The Gemini app now has approximately 950 million monthly active users, according to Google.
The number places Gemini among the most widely used consumer AI products in the world and brings the service closer to the scale of Google’s established products.
Google said daily active users of the Gemini app have tripled during the past year. The company attributes some of this growth to new agentic and personalized features that allow Gemini to perform more complex tasks rather than simply answer individual questions.
The company has been integrating Gemini across Search, Google Workspace, Android, YouTube, Google Cloud and other products. This distribution gives Google an important advantage because users can encounter its AI features without installing a completely separate application.
Gemini is also becoming more closely connected to existing Google accounts, personal information and productivity tools, allowing the assistant to provide answers based on a user’s documents, email, calendar and other permitted sources.
Google AI Mode Surpasses One Billion Users
Google’s AI growth is not limited to the Gemini application.
The company said AI Mode in Google Search has surpassed one billion monthly active users following its international expansion.
AI Mode provides a more conversational search experience that can generate detailed answers, organize information and help users complete multi-step research tasks.
Google has also been combining AI Mode with AI Overviews, its system for displaying generated answers directly above traditional search results.
The company says these AI experiences are increasing overall search activity rather than simply replacing traditional queries. Google also claims that its AI-powered search features continue to send billions of visits to websites every week.
That last point is especially important for publishers and website owners. Many businesses remain concerned that generated answers could reduce the number of people clicking external search results.
Google’s long-term challenge will be demonstrating that AI-powered search can provide direct answers while continuing to support the websites and publishers that supply much of the information used by the search ecosystem.
More Than Nine Million Developers Use Google AI Models
Google reported that more than nine million developers now build with its AI models every month through its APIs and developer products.
Usage across Google’s model APIs has also grown rapidly. The company says its systems are processing approximately 22 billion tokens per minute, up from 16 billion tokens per minute during the previous quarter.
Tokens are the small units of text, code and other information processed by AI models. The number of tokens processed is one way to measure how frequently developers and businesses are using a model.
This growth suggests that Gemini is expanding beyond consumer conversations and becoming part of software applications, automated workflows and enterprise systems.
Developers can use Gemini models for tasks including:
- Generating and reviewing software code.
- Searching and summarizing documents.
- Building customer support assistants.
- Analyzing images, audio and video.
- Creating autonomous agents.
- Automating administrative processes.
- Detecting cybersecurity threats.
- Extracting structured information from large datasets.
Google Cloud Revenue Grows 82%
Google Cloud was one of the strongest areas in Alphabet’s latest financial results.
Cloud revenue reached approximately $24.8 billion during the second quarter of 2026, an increase of 82% compared with the same quarter in 2025.
Google said the growth was driven by demand for enterprise AI infrastructure, Gemini-powered applications and its core cloud computing services.
The operating income generated by Google Cloud also increased from approximately $2.8 billion to $8.8 billion.
These results are important because building artificial intelligence infrastructure is extremely expensive. AI companies require specialized processors, networking equipment, data centers, cooling systems and enormous amounts of electricity.
Google owns much of the technology behind its AI stack, including custom Tensor Processing Units, data center infrastructure, cloud services and the Gemini model family.
This allows the company to provide businesses with an integrated platform instead of depending entirely on hardware and software supplied by other companies.
Gemini Enterprise Expands Across Large Companies
Google said nearly 90% of Fortune 100 companies are now using Gemini Enterprise.
Gemini Enterprise provides tools for organizations to create agents, connect AI systems to internal data and automate business processes.
The company highlighted customers using the technology for analytics, wealth management, customer engagement, commerce and knowledge management.
Google’s Agent Development Kit, a framework for building and deploying enterprise agents, has also reached nearly 70 million total downloads.
Enterprise adoption may become one of the most important areas of the AI market. Consumer assistants attract significant attention, but business customers are more likely to pay for large volumes of tokens, private infrastructure, security controls and long-term integrations.
Companies also tend to remain with a cloud provider after connecting their databases, applications and internal processes to its platform.
Google Begins Training Gemini 4
Google confirmed that Gemini 3.5 Pro is currently being tested and that development of the next major model generation is already underway.
According to Sundar Pichai, Google has started its most ambitious pre-training run yet for Gemini 4.
Pre-training is the stage during which an AI model learns patterns from enormous collections of text, code, images and other data.
Training a frontier model can require thousands of specialized processors operating for extended periods. The process is expensive, technically complex and difficult to restart if something goes wrong.
Google did not provide a release date or detailed specifications for Gemini 4.
The announcement should therefore be viewed as confirmation that development is underway rather than confirmation that the model will be available soon.
The company is currently competing with models from OpenAI, Anthropic and several Chinese AI laboratories. Each new model generation is expected to improve reasoning, coding, multimodal understanding and the ability to operate tools autonomously.
Google Is Spending Heavily on AI Infrastructure
The rapid expansion of Google’s AI products is requiring enormous infrastructure investment.
Alphabet spent approximately $44.9 billion on property and equipment during the second quarter. The company reported negative quarterly free cash flow of approximately $5.9 billion after these investments.
Alphabet also raised nearly $50 billion in net proceeds through an equity offering during June. The company said the funds would support general corporate purposes, including the expansion of AI infrastructure and global computing capacity.
This does not necessarily mean Google’s core business is losing money. The company generated approximately $39.1 billion in operating cash during the quarter.
However, spending on data centers and computing equipment exceeded that amount.
The figures illustrate the scale of the AI infrastructure race. Even one of the world’s largest technology companies must invest tens of billions of dollars to continue expanding its models and services.
Alphabet Reports $119.8 Billion in Revenue
Alphabet’s total quarterly revenue increased by 24% to approximately $119.8 billion.
Google Search and related advertising revenue grew by 17%, while YouTube advertising revenue increased by 13%.
The company reported net income available to common shareholders of approximately $112.1 billion. However, most of the increase came from gains related to Alphabet’s equity investments rather than its normal operating activities.
The earnings release reported approximately $99 billion in gains from equity securities during the quarter.
This distinction is important because those investment gains may fluctuate significantly and should not be interpreted as recurring profit from advertising, cloud computing or AI products.
Why These Numbers Matter
Generative AI companies frequently announce impressive demonstrations and benchmark results, but user adoption and revenue provide another way to measure progress.
Google’s latest numbers show that Gemini is gaining users at a large scale and that businesses are spending more on the infrastructure required to build and operate AI systems.
The company benefits from several advantages:
- A large existing user base across Search, Android and Workspace.
- Global cloud computing infrastructure.
- Custom AI processors.
- Access to large enterprise customers.
- Strong research teams at Google DeepMind.
- The ability to distribute Gemini through existing products.
However, Google also faces significant challenges.
The company must control infrastructure costs, maintain the quality of Search, protect publisher traffic and release models that remain competitive with OpenAI and Anthropic.
It must also convince businesses that Gemini agents can operate securely with sensitive company data.
What Happens Next?
The next major developments to watch will include the wider availability of Gemini 3.5 Pro, additional enterprise agent features and new information about Gemini 4.
Investors will also monitor whether the rapid growth of Google Cloud can justify Alphabet’s increasing infrastructure expenses.
For consumers, the most noticeable change may be the continued integration of Gemini into products they already use.
Google is attempting to make artificial intelligence a standard layer across Search, productivity software, mobile devices, video and cloud computing.
With 950 million monthly Gemini users and more than one billion AI Mode users, that strategy is already reaching an enormous audience.
Final Thoughts
Google’s second-quarter results provide one of the clearest pictures yet of how quickly artificial intelligence is moving from experimentation into everyday products and business infrastructure.
Gemini is approaching one billion monthly users, millions of developers are building with Google’s models and enterprise demand is driving exceptional growth in Google Cloud.
At the same time, the company is spending tens of billions of dollars to support that expansion.
The AI race is therefore becoming a test of both intelligence and infrastructure. Building a powerful model is only part of the challenge. Companies must also provide enough computing capacity to serve hundreds of millions of users reliably and affordably.
Sources
- Alphabet second-quarter 2026 financial results.
- Sundar Pichai’s remarks from Alphabet’s Q2 2026 earnings call.
- Alphabet filing with the U.S. Securities and Exchange Commission.
- Associated Press coverage of Alphabet’s quarterly results.