One Sentence, and AI Builds Your Whole Creative Workflow
I put a node-based creative canvas in the browser: describe your goal, AI generates the workflow — storyboards, image-to-video, one-click sharing. Free to play.
Anyone who has used ComfyUI knows how powerful node workflows are — and how brutal the learning curve is: environments, wiring, parameters, all before your first image.
So I built the opposite, right here at /flow: you don't learn the tool — AI builds the workflow for you.
One sentence → a full workflow
Open the canvas and type a goal, like:
Split "a late-night convenience store encounter" into a 4-panel storyboard, film photography style
AI plans the whole thing: four parallel prompt → image chains feeding one result node, with the character description repeated verbatim in every shot — the key to keeping faces consistent across panels. Hit Run as-is, or tweak any node first.
Pay only for what you change
My favorite design: nodes that already have results are skipped on the next run. Changed panel two's description? Re-running regenerates only that panel. Want to re-roll a single node? Hit ▶ on it. Time and quota go only where you actually made changes.
From one image to a clip
An image node's output plugs straight into a video node as the first frame; connect two images and you get a keyframe transition. The free model runs the whole pipeline; switch to premium models for 2K/4K output and stronger consistency — the canvas options adapt to each model's capabilities.
Share what you build
"Share" creates a read-only link anyone can open — and a "copy to my canvas" button lets them build on your workflow. The workflow itself becomes shareable content.
Free to sign up, with daily free quota. Start at /flow, or warm up in chat mode at /tools.