US-based Alphabet Inc.-backed technology firm Google CEO Sundar Pichai revealed that over 8.5 million developers are currently utilizing its AI models monthly. The company anticipates a capital expenditure of nearly $190 billion this year, primarily focusing on AI data centers, custom silicon, and model training.
Pichai noted a significant surge in demand for Google’s AI ecosystem among developers, enterprises, and consumers. Google’s model APIs are processing close to 19 billion tokens per minute, with more than 375 Google Cloud customers handling over one trillion tokens each in the past year.
Emphasizing the increasing adoption of Google’s AI products, Pichai mentioned that the Gemini app has surpassed 900 million monthly active users, doubling from the previous year. AI-powered search features like AI Overviews have reached over 2.5 billion monthly users, while AI Mode has exceeded one billion monthly active users.
Google also introduced new AI tools and enhancements, such as Gemini 3.5 Flash, a faster AI model concentrating on coding and real-world workflows, and Gemini Spark, a personal AI agent for user task execution. The company updated its AI infrastructure with the latest Tensor Processing Units (TPUs) to facilitate faster AI training and inference at scale.
Remaining committed to an AI-first approach, Pichai described AI as the most transformative technology for advancing Google’s mission and enhancing lives globally. Google unveiled ‘Ask YouTube,’ an AI-powered feature aiding users in discovering relevant video moments through conversational queries, and ‘Docs Live,’ enabling document creation and organization via voice commands.
Furthermore, Google expanded its AI transparency efforts through SynthID, a watermarking technology for AI-generated content. Several companies, including OpenAI, Kakao, and Eleven Labs, are adopting this technology to enhance the verification of AI-generated images, videos, and audio.
