Google just launched Gemini Omni, a new AI model that can create videos from text, images, audio, or any combination of the above. It’s already available for free on YouTube Shorts, and subscribers can use it today in the Gemini app.
The Gist
- Gemini Omni Flash is live now for Google AI subscribers and free on YouTube Shorts and YouTube Create.
- It edits videos through natural conversation and understands physics, motion, and real-world behavior.
- Gemini 3.5 Flash also launched as the new default AI brain behind Google Search globally.
What Gemini Omni Actually Does
For a while, creating videos with AI meant working with rigid templates or hoping the result looked halfway decent. Gemini Omni changes the equation. You feed it anything: a photo, a voice clip, a piece of text, or a combination of all three. It produces a video. That’s the short version.
The longer version is more impressive. Gemini Omni doesn’t just stitch images together. It understands how objects move, how gravity works, how light behaves. Ask it to show a coffee cup falling off a table, and the physics will look right. Ask it to apply a visual style from one video to another, and it will do it frame by frame through a simple chat conversation.
Google calls this “conversational video editing.” You start with an idea, generate something, then keep refining it by typing. “Make the sunset warmer.” “Slow down the second half.” “Add rain.” The model builds on each previous instruction rather than starting from scratch each time. This is new behavior for AI video tools.
There is also an avatar feature. You provide your voice, and the model creates a personalized digital avatar that can appear in your videos. All videos generated by Gemini Omni carry a SynthID watermark, an invisible digital tag that marks AI-generated content. Google built this in from day one.
The first model in the Omni family, Gemini Omni Flash, is the version available right now. Google has indicated that future Omni models will expand beyond video to generate audio and other output types. For now, video is where it starts.
How to Use It Today (And It’s Free)
You don’t need to pay anything to try Gemini Omni. Google rolled it out for free on YouTube Shorts and the YouTube Create App starting this week. If you already use YouTube to post videos, you’ll find it inside the creation tools. No separate signup, no waiting list.
If you’re a Google AI subscriber (Plus, Pro, or Ultra plans), you can access Gemini Omni directly in the Gemini app and in Google Flow, Google’s AI-powered creative workspace. The difference from the free version is mostly about length, quality settings, and the number of videos you can generate per month.
For developers who want to build products on top of it, the API access is coming in the next few weeks. Enterprise customers will also get access through that same API rollout. As AI PCs like the Nvidia RTX Spark-powered devices arrive later this year, tools like Gemini Omni are the kind of software they’re designed to run.
The practical use cases are already obvious. Short social media videos. Product demos. Explainer clips. Travel recaps from photos. Anyone who has ever paid a video editor for a 30-second clip now has a free alternative that works from a single text prompt.
The big limitation right now is the output. Gemini Omni starts with video generation and will expand over time. If you need high-resolution, long-form, broadcast-quality content, this is not there yet. But for short clips, social content, and personal projects, it works today.
What Gemini 3.5 Flash Means for Google Search
Alongside Omni, Google launched something just as significant for everyday users: Gemini 3.5 Flash. This is a new AI model that is now the default engine powering Google Search’s AI Mode globally. If you have used AI Mode in Search, it just got a major upgrade without you doing anything.
Gemini 3.5 Flash is described as combining “frontier intelligence with action.” In plain terms: it’s faster than the previous model, better at understanding complex questions, and significantly improved at coding tasks. Google says it outperforms the previous generation, Gemini 3.1 Pro, on almost all benchmarks.
The model is also available to developers immediately through the Gemini API and AI Studio. Google is already working on Gemini 3.5 Pro, which is currently in internal testing and expected to be released next month. The jump from 3.1 to 3.5 is notable because it skips a version number entirely, signaling a more substantial leap than a typical incremental release.
What does this mean for someone who just uses Google Search? In short: the AI answers you get when you ask complex questions should be more accurate, more useful, and faster. The “AI Mode” feature, which was already rolling out globally, is now running on a smarter engine without any change on your end.
Google also announced AI Search agents at the same event. These are background tools that can research topics on your behalf around the clock. The combination of Gemini 3.5 Flash and these agents represents Google’s clearest statement yet that AI is no longer just a search box feature. It’s the new infrastructure of how you find information online.
Follow the story on AI Noobies.
About the Author
