Build AI workflows on a visual canvas

Wire text, image, video and audio models into a node graph — or let the AI assistant build and connect it for you. Runs in the browser, no download, shareable with a link.

Workflow — Untitled
Prompt

“a leopard cub in a misty jungle, cinematic”

Image
first frame
Seedance 2.0
9:16 · 8s · rendering

What you can build

Batch ad variations

Fan one graph out into dozens of ad variants in a single run using the AI generator

Image-to-video pipelines

Chain a first-frame image straight into Seedance 2.0

Voiceover + captions

Add generated narration and sound effects from AI audio

Render and export

Composite the graph output and export finished MP4s in the timeline editor

Remake winning ads

Feed library ads and your product into Ad Remake

Reusable templates

Save a graph as a template and re-run it with new inputs in the workflow editor

Frequently asked questions

What is Planegraph's workflow editor?+

Planegraph's workflow editor is a visual, node-based canvas for generative AI. You connect model nodes — text-to-image, image-to-video, voice, lipsync, compositing — into a graph and run the whole graph as one automated pipeline, similar to n8n but for AI media. It runs entirely in the browser: nothing to download or host.

How is Planegraph different from ComfyUI?+

Planegraph runs entirely in the browser — no local install, no GPU, and no self-hosting, where ComfyUI is a locally hosted node tool you download and run yourself. Planegraph also has an AI assistant that builds large workflows for you, auto-placing nodes and connecting edges, and every workflow can be shared with a link that opens directly in the browser.

Do I need to write code to build a workflow?+

No. The workflow editor is fully no-code: you add nodes, connect them, and run. The AI assistant can also build the graph for you from a plain-text brief — adding the nodes, auto-placing them on the canvas, and connecting the edges.

Can the assistant build a workflow for me?+

Yes. Describe the pipeline in plain text — for example "4-scene product ad, image to video, with voiceover" — and the assistant adds the nodes, auto-places them, and wires the connections. It handles large graphs with dozens of nodes.

Can I share a workflow?+

Yes. Every workflow has a shareable link. Anyone you send it to opens the graph in their browser — no download or local setup required.

Which AI models can I chain in a workflow?+

You can chain Seedance 2.0, Veo 3.1, Nano Banana 2, GPT 1.5 Image, ElevenLabs and more in a single graph, mixing image, video and audio models. New models are added as they are released.

Can I generate ad variations in bulk?+

Yes. A workflow can fan a single graph out into many variations in one run — different products, copy, formats or models — which makes it well suited to batch ad generation at scale.

Is the workflow editor free?+

Yes, the workflow editor is free to start with no credit card required, and you keep commercial usage rights to what you create. New users also get 500 free credits for Seedance 2.0, Veo 3.1, Nano Banana 2 and more.