Runway launches node-based Workflows for building custom AI creative pipelines

Runway has introduced Workflows, a node-based system within its platform that lets users build custom creative pipelines by connecting multiple AI models, modalities, and generation steps. The feature works similarly to other AI canvas tools like FLORA, giving creators greater control over outputs and allowing for more personalized automation than standard generation tools.

Each node represents a specific function or generative model, while links transfer data between inputs and outputs. Users can combine and experiment with nodes to create reusable templates or varied results. Color coding with orange for text, pink for image, and green for video helps organize complex setups more intuitively.

Workflows include three main node types: Input nodes for text or media uploads, Media model nodes for image and video generation, and LLM nodes for text-based automation and prompt refinement. The Workflow Editor supports adjusting parameters like aspect ratio, resolution, and temperature, and users can run entire workflows or individual nodes for testing. The feature is currently available to Creative Partners and Enterprise customers, with broader access coming soon.