Gradio Launches gr.Workflow: Visual Pipeline Builder with Auto-Deploy to REST API and Hugging Face Spaces

Gradio has released gr.Workflow, a new feature that lets developers construct multi-step generative media pipelines as visual, drag-and-drop node graphs directly within the Gradio framework. Pipelines can chain together image generation via FLUX, background removal, text-to-speech, video animation with LTX-Video, LLM calls, and other AI models — with every intermediate result visible at each node during execution.

Each workflow automatically exposes its outputs as individual REST API endpoints and can be deployed to Hugging Face Spaces with a single command, requiring no additional configuration. The release includes several live demo pipelines — including an AI Media Studio, a fan-out generative art tool, and a ZeroGPU-powered image animator — illustrating patterns such as sequential chaining, parallel fan-out, and GPU-accelerated local model execution.