Alphabet (GOOG) Opening Keynote summary

AI innovation and infrastructure advancements

• Over 3,000 product advances shipped in 2024, with Google Cloud expanding to 42 regions and new subsea cables announced for global connectivity.

• $75 billion CapEx planned for 2025, focused on servers, data centers, and AI compute to support global services.

• Launch of Cloud WAN, offering 40% faster performance and up to 40% lower total cost of ownership for enterprises.

• Introduction of Ironwood, the seventh-generation TPU, delivering 3,600x performance improvement and 29x energy efficiency since the first TPU.

• Quantum chip Willow achieves exponential error reduction, paving the way for scalable quantum computing.

AI models, products, and customer impact

• Gemini 2.5, the most advanced AI model, leads industry benchmarks and is available across Google products and platforms.

• Gemini 2.5 Flash offers low-latency, cost-efficient reasoning, allowing users to balance performance and budget.

• AI models like Veo 2, Imagen 3, Chirp 3, and Lyria enable generative media across video, image, audio, and music, available on Vertex AI.

• Customers such as McDonald's, Honeywell, and Intuit leverage AI for operational transformation, predictive maintenance, and data-driven insights.

• Vertex AI and Gemini usage has grown over 40x, with billions of API calls monthly and over 200 curated models accessible.

Multi-agent systems and enterprise AI adoption

• New agent development kit and Google Agent Space enable building, managing, and deploying multi-agent systems at scale.

• Agent Space integrates with Chrome, allowing enterprise data access and AI-powered agents directly from the browser.

• Customer Engagement Suite introduces next-gen features: human-like voices, emotion comprehension, streaming video support, and no-code agent building.

• Industry-specific agents drive impact in sectors like retail, healthcare, finance, and creative industries, with partners such as Reddit, Mercado Libre, and Verizon.

• Data agents, coding agents, and security agents streamline workflows, automate analysis, and enhance security for organizations like Mattel, Spotify, and Charles Schwab.