With the rapid adoption of AI models in software products, Chrome’s introduction of local-first, on-device AI models offers developers a powerful new way to build intelligent and fast AI features directly within the browser. This talk will explore Chrome’s Built-in AI models, their architecture, APIs, developer tooling, and also demonstrate how it supports low-latency AI experiences without relying solely on cloud-deployed models.

We’ll walk through the architecture behind Chrome’s on-device models, discuss real-world edge cases and common concerns from integrating the built-in AI APIs, and outline key system and compatibility requirements that every developer should know. This talk will also mention the evolution of Chrome’s AI ecosystem, from its early releases to current capabilities, along with a preview of upcoming features and insights from its developer community.

For advanced listeners in the audience, this talk will include Chrome’s hybrid AI support pairing with Firebase to extend applications with cloud-backed storage and processing, enabling developers to blend on-device intelligence with online workflows.

To keep this talk interactive and practical, I will use my open-source Chrome extension project built during the recently concluded Chrome AI hackathon as a demo, highlighting how the Prompt and Summariser APIs are being used in the Chrome extension.

Web AI libraries and models crossed over 1 billion downloads in the last 2 years alone. If you're not yet using machine learning on the client side, now is the time to see what's possible and the benefits it can bring your business over cloud based solutions - such as lower costs, privacy for your end users, and low latency responses suitable for real-time applications.

Learn what's new with the state of Web AI at Google and discover new generative AI models you can run client-side in the browser entirely offline along with key updates from Chrome, TensorFlow.js, MediaPipe, Core ML, Visual Blocks, and beyond.

A must see talk for anyone working in the web space (or is using JavaScript) so you can bring your AI powered ideas to life faster than ever before.

AI interfaces are starting to outgrow the chatbox. Instead of returning only text, agents can assemble the interface a user needs in real time. The challenge is enabling that flexibility without giving an AI model unrestricted control over the frontend.

This talk presents a bounded architecture for generative UI using React and Google’s A2UI protocol. We’ll explore how agents can stream declarative interface schemas that are validated and mapped onto trusted components from an existing design system. We’ll also cover component contracts, server actions, client-side state, streaming, and the boundary between agent-generated structure and application-controlled behaviour.

Performant Architecture for a Fast Gen AI User Experience

More and more developers are being asked to build Generative AI experiences into their web applications everywhere. However, the user's experience with Generative AI can sometimes be slow and frustrating. It’s been years since we would gladly wait over 10 seconds for a response to anything online. As AI processing gets faster, there are new limiting factors that could be negatively affecting your AI user experience.

In this talk, we will cover different ways to minimise laggy experiences in web applications when interacting with third-party Generative AI tools. We’ll be using a practical demonstration to show how you can reduce latency as we build one of my favourite sci fi gadgets, the babel fish.