Google launches Gemini 3.5 Flash and Gemini Omni AI models

Google Just Leveled Up: Meet Gemini 3.5 Flash and the Mind-Blowing Gemini Omni

Google’s I/O conference is always a bit of a whirlwind, but this year, the spotlight was firmly on the next generation of intelligence. Google officially pulled back the curtain on the Gemini 3.5 family and a brand-new video powerhouse called Gemini Omni. Whether you’re a developer, a creator, or just someone who uses Google Search every day, these updates are about to change how you interact with your devices.

Gemini 3.5 Flash: Speed Meets Serious Brainpower

First up is Gemini 3.5 Flash. If you’ve felt like AI models were getting smarter but a bit sluggish, this is the answer. It’s now the default model for everyone using the Gemini app and AI Mode in Google Search. What makes it special? It hits that sweet spot where you get the ‘pro’ level intelligence of a flagship model but at the lightning-fast speeds the Flash series is known for.

For the tech-savvy crowd, this is now Google’s strongest model for coding and ‘agentic’ tasks—meaning it’s better at actually performing multi-step actions rather than just generating text. It even edges out Gemini 3.1 Pro on complex benchmarks, particularly when it comes to multimodal understanding. In short: it’s fast, it’s smart, and it’s already live.

Gemini Omni: Turning Any Prompt Into High-End Video

The real showstopper, however, is Gemini Omni. This isn’t just another text-to-video tool. Omni allows you to mix and match inputs—think images, audio clips, video snippets, and text—to generate high-quality video that actually understands the real world.

One of the coolest features is the conversational editing. Once you’ve generated a video, you don’t have to start over if something isn’t quite right. You can simply talk to the AI to refine the scene, swap out details, or change the lighting, all while the model keeps the original ‘thread’ of the scene intact. It’s like having a digital film editor at your beck and call.

Real-World Physics and Digital Avatars

One of the biggest hurdles for AI video has always been ‘the uncanny valley’—things looking just a bit too weird. Gemini Omni Flash addresses this with a much deeper understanding of physics. We’re talking about gravity, kinetic energy, and how fluids move. This means scenes look more grounded and realistic than ever before.

The update also introduces personal Avatars and voice cloning. You can essentially create a high-fidelity digital version of yourself to star in these videos. And for those worried about authenticity, Google is leaning heavily into safety; every video generated includes SynthID digital watermarking to ensure AI-generated content is easy to identify.

Where Can You Get It?

If you’re a subscriber to Google AI Plus, Pro, or Ultra, you can start playing with Gemini Omni Flash today via the Gemini app and Google Flow. But the best news for casual creators? It’s also rolling out as a free tool for users on YouTube Shorts and YouTube Create. The barrier to entry for high-quality video production just got a whole lot lower.