Case Study: OpenArt scales custom AI image workflows on hundreds of GPUs with Modal

How OpenArt scaled their Gen AI art platform on hundreds of GPUs

OpenArt, a platform for AI image generation with millions of users, needed to deploy its proprietary ComfyUI workflows but found API providers too inflexible and managing its own infrastructure on AWS or GCP to be a slow, hours-long process. This deployment bottleneck hindered the startup's ability to scale its sophisticated image editing tools for its growing customer base.

By using Modal's programmatic infrastructure, OpenArt was able to deploy its custom ComfyUI workflows in minutes instead of hours. Modal’s solution provided the necessary customizability and autoscaling, which allowed OpenArt to power over 100 workflows and seamlessly scale to hundreds of GPU containers to serve its millions of users without overloading its tech stack.

OpenArt

Coco Mao

CEO and Co-founder