Google Is Merging Its AI Creative Tools Into Flow. It
Google announced that Whisk and ImageFX are moving into Flow, creating a unified workspace for AI image and video generation. Users have created over 1.5 billion images and videos on the platform since launch.
Google has a tool problem. Not a lack of tools. Too many of them.
Over the past year, the company launched Flow for video generation, Whisk for image remixing, ImageFX for text-to-image, and Nano Banana for high-fidelity images. Each one lived in its own app with its own interface. If you wanted to generate an image in ImageFX and then animate it into a video in Flow, you had to download the image, switch apps, and upload it again. It was clunky and honestly kind of embarrassing for a company that preaches integration.
That's changing. Google announced this week that Whisk and ImageFX are moving directly into Flow. Everything lives in one workspace now. Generate an image with Nano Banana, use it as a frame for a Veo video, and edit both without leaving the app.
What Changed
The biggest upgrade is having Nano Banana, Google's latest image generation model, built into the core Flow experience. Previously, Flow was mostly a video tool. Now it handles the full pipeline from static image to animated video.
Here's what that looks like in practice. You type a prompt, get a high-fidelity image, then with one click turn that image into a video clip. Want to change something in the image first? Use the new lasso tool to select a region and describe what you want different. "Remove the man." "Add koi fish in the water." You can even draw directly on images to guide the AI.
The video editing features got upgrades too. You can extend clip length, add or remove objects from video, and control camera motion with pans and zooms. These aren't new capabilities individually, but having them all in one place matters. Creative workflows break down when you have to switch between five different tools for five different steps.
Google says users have created over 1.5 billion images and videos on Flow since it launched. That's a big number, though the company didn't break down how many were images versus videos, or how many were from paying users versus free accounts. Image generation in Flow is now free, which suggests Google is prioritizing adoption over monetization for now.
The Migration Question
Starting in March, existing Whisk and ImageFX users can opt in to transfer all their projects and assets into Flow. This is important because creative tools live and die on user libraries. Asking people to start over in a new app is a great way to lose them. Google seems to understand this. The migration preserves everything.
But there's a risk. Whisk and ImageFX attracted specific audiences who liked those tools for specific reasons. Whisk users loved the remix workflow. ImageFX users liked the simplicity. Folding everything into Flow means those users get new features but also new complexity. A unified workspace sounds great in a blog post. In practice, it can feel overwhelming if the interface isn't clean.
Google addressed this partly with flexible asset management. A new grid view lets you search, filter, and sort across images and videos. You can group assets into collections. There's drag-and-drop for organization. And you can reference specific assets using an @ symbol, which is a nice touch that makes the workspace feel more like a conversation than a file browser.
Where Flow Stands Against the Competition
Google's real competition here isn't Canva or Adobe, though those companies should pay attention. It's the collection of AI video tools that sprouted up over the past year. Runway Gen-3, Pika, Kling, Seedance from ByteDance, and others. Each one handles a piece of the creative pipeline. None of them do everything.
Flow's advantage is integration. When you control the image model, the video model, and the editing tools, you can pass context between them in ways that separate tools can't. A video generated from an image inherits the style and composition of that image automatically. You don't lose fidelity in translation.
The disadvantage is that Google moves slowly on consumer products. Flow has been in preview for months and still isn't available everywhere. Meanwhile, competitors ship updates weekly. Runway pushes new features at a pace that makes Google look like a government agency.
What This Means for Creators
For anyone doing creative work with AI tools, this is good news regardless of which platform you use. The trend is toward unified workspaces, and when Google moves, competitors follow. Expect Runway, Pika, and the rest to announce their own integrated creative suites within months.
The free image generation in Flow is also worth noting. Google can afford to give away AI-generated images because it's subsidizing the cost against its cloud revenue. Smaller competitors can't match that pricing, which will pressure the entire market toward lower prices or free tiers.
For professional creators, the real test is output quality and control. Can Flow produce images and videos that are good enough for commercial work? The lasso tool and camera controls suggest Google is targeting professionals, not just hobbyists. But the proof is in the output, and Google's track record on creative tools is mixed at best.
One last thing. The 1.5 billion creations number hints at something bigger. If Google can build a massive library of user-created content and preference data, it gets a feedback loop for training better models. Every edit, every regeneration, every prompt refinement teaches the model what users actually want. That data advantage compounds over time.
The creative AI tool market is consolidating. Google just made a big move toward owning the full pipeline. Whether it executes well enough to hold that position is another question entirely.
Get AI news in your inbox
Daily digest of what matters in AI.
Key Terms Explained
A mechanism that lets neural networks focus on the most relevant parts of their input when producing output.
Contrastive Language-Image Pre-training.
AI models that generate images from text descriptions.
The process of teaching an AI model by exposing it to data and adjusting its parameters to minimize errors.