At a glance
Google Flow is Google’s AI filmmaking app, and at I/O 2026 it gained a Flow Agent: a creative partner that can plan a project, generate scenes, and edit your timeline through plain conversation. It runs on Gemini Omni, Google’s new video model that keeps characters and voices consistent from shot to shot. Together they turn “make me a video” into a back-and-forth with an assistant instead of a pile of separate prompts. Here is what it does, who it is for, and where it still falls short.
If you have ever wanted to make a short video but found AI tools fiddly, prompt one clip, prompt another, then stitch them together by hand, Google’s update to Flow is aimed straight at you. Flow now has an agent that handles the multi-step work for you, and a new model, Gemini Omni, that edits video the way you would talk to a human editor. Here is the plain-English version of what changed and whether it is worth your time.
What is Google Flow?
Google Flow is an AI filmmaking tool from Google Labs. You describe a scene, and Flow generates video for it using Google’s Veo model, pulls in images from its Nano Banana image engine, and lets you arrange clips on a timeline. It launched in 2025 as a way to make short, AI-generated films without a camera or editing software. The 2026 update is what makes it interesting for non-experts, because you no longer have to drive every step by hand. For the wider picture of where these tools stand, see our look at the real state of AI video in 2026.
What is the new Google Flow agent?
This is the headline change. The Flow Agent is a creative partner built into Flow that can plan and reason through a whole project, not just make one clip. Google describes it as a tool that helps with early brainstorming, creating, and editing, all under your control. In practice that means you can:
- Ask it to batch-generate several versions of a scene at once.
- Get plot and dialogue suggestions instead of staring at a blank page.
- Keep refining the same project by talking to it, with no manual re-prompting each time.
So instead of you breaking a video into a dozen separate prompts, the agent handles the multi-step path and checks in with you along the way.
What is Gemini Omni, and how does it edit video?
Gemini Omni is the new Google model that powers Flow’s editing. Google’s pitch is that it can create or change video from almost any input, and its standout trick is consistency: it keeps a character’s face and voice the same across every scene, which has been the hardest problem in AI video. You can edit by talking to it, for example “change the lighting to golden hour” or “make this cut faster,” and because it understands your whole timeline, you can say “match the color of the opening shot” and it knows what you mean. If you want the deeper dive on the model itself, we covered Gemini Omni when it launched and how to prompt it well.
What does the agentic video experience actually look like?
Put the two together and the whole experience changes shape. A rough session looks like this:
- You tell the agent the idea: “a 30-second ad for a coffee cart, warm and friendly.”
- It proposes a plan: the scenes, the shots, a line of voiceover.
- It generates the scenes with Veo, keeping your character consistent through Omni.
- You give notes in plain language, and it edits the timeline rather than starting over.
- You repeat until it is right, then export.
That is the “agentic” part: the AI carries the multi-step load, and you stay the director making the calls.
Who is Google Flow for, and what does it cost?
Flow is aimed at anyone who wants to make video without a production crew: creators, marketers, teachers, and hobbyists. The new Gemini Omni editing is available inside Flow to Google AI subscribers worldwide, and Google has released a Flow mobile app in Android beta (18 and over), with an iPhone version planned later. If you already pay for a Google AI plan, it is included. If not, it sits behind that subscription. For most beginners, the web version is the place to start, and our roundup of the best AI tools for beginners can help you decide what else is worth your time.
What can Google Flow not do yet?
Plenty, and it helps to be clear-eyed. AI video still struggles with longer, complex stories, fine detail, and anything that needs exact real-world accuracy. The agent can plan and generate, but it cannot supply the idea, the taste, or the reason a video is worth making in the first place. It is a strong assistant for the mechanical parts of filmmaking, not a replacement for the person with something to say. Used well, it frees you to spend your time on the creative calls that actually matter.
The Beginners in AI take: Flow’s agent is the clearest sign yet of where AI video is heading, from “type a prompt, get a clip” to “describe a project, shape it with an assistant.” That is a real step up for beginners, because the hardest part was never the idea, it was the dozen fiddly steps in between. Let the agent carry those. Keep the vision yours.
Two ways to go further
The AI Prompt Library
1,000+ ready-to-use prompts for Claude, ChatGPT, and Gemini. Stop staring at a blank box.
Get it for $39 →1-on-1 Custom AI Tutorial
A private, beginner-friendly session on the AI tools you choose, built around your goals.
Book for $99 →Get Smarter About AI Every Morning
Free daily newsletter. Built for people who want to use AI well, not chase every model.
Free forever. Unsubscribe anytime.
Common questions about Google Flow
Is Google Flow free?
There is free access to the basic tool through Google Labs, but the new Gemini Omni editing and the Flow Agent sit behind a Google AI subscription. Check your plan before counting on those features.
What model does Google Flow use?
Flow generates video with Google’s Veo model and now edits with Gemini Omni, plus the Nano Banana engine for images. The Flow Agent is what ties them together into one workflow.
Can Google Flow edit a video I already filmed?
Flow is built mainly to generate and assemble AI video on its own timeline. Gemini Omni adds strong editing of that footage by conversation, but it is not a traditional editor for the clips on your phone.
Is there a Google Flow app?
Yes. Google released a Flow app in Android beta for users 18 and older, with an iPhone version planned later. The web version is available now and is the easiest place to begin.
Sources
You may also like
- Gemini Omni: Google’s video leap
- Gemini Omni video prompting
- The real state of AI video in 2026
- What happened to Sora 2
- How to use Gemini
- New to AI? Start here
Two ways to go further
The AI Prompt Library
1,000+ ready-to-use prompts for Claude, ChatGPT, and Gemini. Stop staring at a blank box.
Get it for $39 →2-Hour Live AI Crash Course
A private, beginner-friendly session across Claude, ChatGPT, Gemini, and the wider landscape.
Book for $125 →