Client side Web AI agents to build smarter future user experiences

In a world where AI agents and humans coexist, what does the web of the future look like, and how could it work? What are agents, and how can Google’s latest client-side gen AI models help you perform agentic behaviors for real-world applications, to complete common tasks on your website faster? We’ll share a vision for an agentic internet capable of using local LLMs to perform actions on a user’s behalf through natural interfaces like text or voice, entirely in the JavaScript stack, along with the emerging world of hybrid Web AI, that embraces the benefits of Cloud and Web AI together with current generation devices. Resources: Web AI Agent Github → https://goo.gle/web-ai-agent Web AI DJ Github → https://goo.gle/web-ai-dj AI Model Proxy Cache Github → https://goo.gle/ai-model-proxy-cache Learn More about Web AI → https://goo.gle/WebAIVideos Connect with Jason → https://goo.gle/jason-mayes Speaker: Jason Mayes Check out the AI session track from Google I/O 2025 → https://goo.gle/io25-ai-yt Check out all sessions from Google I/O 2025→ https://goo.gle/io25-sessions-yt Subscribe to Chrome for Developers → https://goo.gle/ChromeDevs Event: Google I/O 2025 Products Mentioned: AI/Machine Learning