Example Workflows
June 9, 2026 · View on GitHub
Ready-to-run graphs. Drag the .json onto the ComfyUI canvas (or Workflow → Open),
open the AtlasCloud Client node, and replace PASTE_YOUR_ATLASCLOUD_API_KEY with your
free API key. Then hit Run — no local GPU or model weights needed.
| File | What it does | Models | Nodes |
|---|---|---|---|
01-text-to-image.json | Prompt → image | Nano Banana Pro | 3 |
02-image-to-video.json | Image → ~5s video | Seedance 2 (image-to-video) | 5 |
03-multi-reference-video.json | Up to 8 reference images → video | Seedance 2 (reference-to-video) + Multi Image to Base64 | 7 |
Steps
- Install the nodes — see the root README.
- Get a free API key.
- Open a workflow, click the AtlasCloud Client node, paste your key.
- Run.
Tip: Seedance reference / image-to-video inputs reject real-person photos — use objects or scenes, or switch to a Kling node for people. Image inputs must be ≥ ~768×768.
More example workflows are added as new models land. Want a specific pipeline? Open an issue.