Google turns Flow into an all-in-one image and video studio

If you have ever tried making an AI video, you know the drill: you generate an image in one tool, download it, upload it to a video generator, realize there is a glitch, and jump into an editor to fix it.

Google wants to kill that back-and-forth. On February 25, the company rolled out a major redesign for its creative platform Flow, turning what was once just a prompt box into a unified, all-in-one production canvas.

According to Google, users have created over 1.5 billion images and videos on Flow since its launch last year. But the constant feedback was simple: creators needed far more than an isolated video maker.

Bringing image and video under one roof

The centerpiece of this update is folding Google Labs' standalone experiments, ImageFX and Whisk, directly into Flow. Instead of switching tabs, you can now handle everything in one screen.

The Nano Banana image model is now deeply integrated into the experience. You can generate high-fidelity images for free and immediately route them as visual ingredients or keyframes into the Veo video model without ever leaving the interface.

To keep existing workflows intact, Google says users will be able to migrate all their previous Whisk and ImageFX assets directly into Flow starting in March. The platform also gained an asset grid that lets you filter, search, and group files into project collections.

Lasso tools and natural language fixes

Beyond multimodal unification, Flow is adding fine-grained control features that feel closer to traditional editing suites.

A new lasso tool lets you outline any part of a frame and type natural-language instructions to tweak it, like asking the model to remove a person or add fish into water. You can also sketch directly on the canvas to guide edits.

On the video front, Flow now supports extending existing clips seamlessly, inserting or removing objects across frames, and orchestrating specific camera motion paths. The updated platform is live at flow.google.

Why it matters

We think this is one of Google's smartest product consolidation moves in months.

Generative media has reached a point where raw generation quality is table stakes. The real friction for creative pros has been the chaotic workflow: shuffling files between Midjourney, upscale tools, and video generators. By coupling Nano Banana and Veo inside a single canvas with unified asset management, Flow removes the tedious intermediate file management and makes rapid iteration genuinely feasible.

It also sets up Flow to take on Runway and Pika more aggressively. Instead of keeping ImageFX and Whisk as isolated experiments, Google is finally turning its disparate research toys into a coherent creative suite.

The catch, however, lies in consistency. Natural language inpainting and object removal in video look great in curated demos, but in production, multi-step generative edits often introduce artifacts and break spatial physics. If creators still have to regenerate multiple times to get a clean fix, no amount of interface polish will save them from trial-and-error fatigue.