Turn any ComfyUI workflow into a production-ready API on Runpod

Free to use. Every workflow makes the Wizard smarter.

workflow.json
Load Checkpointsd_xl_base.safetensorsMODELCLIPVAECLIP Text EncodeclipCONDITIONINGa cosmic landscape, glowingbioluminescent flora, 8k...CLIP Text EncodeclipCONDITIONINGblurry, lowres, watermark,jpeg artifacts, ugly...Empty Latent ImageLATENTwidth1024height1024batch_size1KSamplermodelLATENTpositivenegativelatent_imageseed156680208control_after_generaterandomizesteps20cfg8.0sampler_nameeulerschedulernormaldenoise1.00VAE DecodesamplesIMAGEvaeSave Imageimagesfilename_prefixComfyUI
7 nodes1 modelbuilt-in only
200
POST /v2/your-endpoint/runsync
AI-generated cosmic landscape with bioluminescent flora
512 × 512

How It Works

Every workflow we run teaches the Wizard. Coverage compounds with use.

1

Upload

Export your workflow from ComfyUI with Save (not Save (API), which strips the info we need) and drop the JSON here.

2

We resolve

The Wizard identifies every model, custom node, and dependency the workflow needs.

3

We capture

We build the image, run it once, and the Wizard learns each new custom node's schema, so the next workflow that uses it just works.

4

You get a container

Pull the Docker image and run the workflow. No extra setup.

The Wizard's Library

The Wizard catalogues nodes, models, and dependencies from real-world workflows. Whatever ComfyUI setup you throw at it, the Wizard knows how to spin up a working Docker container.

23

Workflows

Verified, ready-to-use Docker images

Growing

Daily

Community contributions expand our coverage

Any

Complexity

From simple to multi-model pipelines

Join the Community

Train the Wizard

Every workflow you submit teaches the Wizard a new ComfyUI trick, and makes it sharper at turning any setup into a production-ready container, no matter how exotic the node graph.