Flutter developers have long relied on server-side solutions to integrate AI into their apps. With Genkit now available as a native Dart SDK, the game has changed — you can build AI-powered features using a single language across your entire stack.

In this talk, we'll explore how Genkit on Dart enables Flutter developers to work with multiple AI providers — Gemini, Claude, ChatGPT, and more — through a unified, type-safe API. We'll cover structured output generation, tool calling, agentic workflows, and the unique middleware system for building intelligent agents with file access, skills, and human-in-the-loop approval.

But the real breakthrough is hybrid AI architecture. With Genkit's Chrome plugin, your Flutter Web apps can run Gemini Nano directly on-device — no server, no latency, no cost — while seamlessly falling back to cloud models for complex tasks. Combined with Genkit's client library for streaming remote flows and a Dart backend deployable to Cloud Run or Firebase, you get a true full-stack AI solution in pure Dart.

We'll walk through practical examples: setting up a Dart AI backend, connecting it to a Flutter app, running on-device inference, and deploying everything to production. Whether you're building chatbots, content generators, or autonomous agents — this talk will show you how to do it all without leaving the Dart ecosystem.