RunComfy

What it does

RunComfy runs ComfyUI in the cloud with no local setup. You drop in a workflow.json and it handles dependencies, custom nodes, and model downloads. It offers on-demand GPUs from 16GB to 141GB VRAM, 200+ ready-to-run templates, and one-click serverless API deployment.

RunComfy is a cloud service for running ComfyUI workflows without a local machine. You drop in a workflow.json, and the auto-setup agent handles dependencies, custom nodes, and model downloads, then gives you a link to run it. The site lists 200+ ready-to-run templates and model-specific pages for FLUX, WAN, LTX, and Qwen Image. Saved workflows can be shared by link or deployed as a serverless API that scales with traffic and scales to zero when idle. The platform offers on-demand GPU machines with VRAM from 16GB to 141GB. Models can be downloaded from Civitai, Hugging Face, and Google Drive, and are persisted across sessions. The homepage states that RunComfy is trusted by 1M+ creators and lists 99.9% uptime and 24/7 support. Pricing details are not shown on the homepage, so check the pricing page before committing.