The next phase of the AI race is no longer just about building larger models, but about making advanced AI faster, cheaper, and widely accessible. Google CEO Sundar Pichai emphasized that efficiency and deployment scale are now as critical as model size, highlighting the company's progress with Gemini 3.5 Flash, which delivers nearly 90% of frontier model performance while being four times faster and significantly more affordable. This shift reflects a broader industry move toward practical AI integration at scale.

Google is processing over 3.2 quadrillion tokens monthly—up from 480 trillion a year ago—with 8.5 million developers actively building on its Gemini platform. Consumer adoption is surging: AI Overviews now reach 2.5 billion users monthly, AI Mode has hit 1 billion users, and the Gemini app has grown to 900 million active users. These numbers underscore how deeply AI is being woven into Google’s ecosystem.

To support this expansion, Google introduced new tools like Gemini Spark, an AI assistant that performs background tasks across Workspace apps, and Gemini Omni, a multimodal model capable of generating video. The company also launched Anti Gravity 2.0, a platform for developing autonomous AI agents. Pichai positioned Search as Google’s most used AI product, noting that queries are becoming longer, conversational, and multimodal.

Google also announced key partnerships to strengthen AI infrastructure and transparency. OpenAI, Nvidia, Kakao, and Lightricks are adopting SynthID, Google’s AI watermarking tech, while Amazon, Meta, Microsoft, Salesforce, and Stripe have joined its Universal Commerce Protocol for AI-powered shopping. The overarching goal, Pichai said, is bringing frontier AI capabilities to billions of users worldwide.