> ## Documentation Index
> Fetch the complete documentation index at: https://docs.imagine.art/llms.txt
> Use this file to discover all available pages before exploring further.
> Build powerful AI pipelines on an infinite visual canvas. Connect 50+ models, organize your workspace, and collaborate with your team.
# Intro to Workflows
Workflows (also known as **Imagine Flow**) is ImagineArt's visual node-based canvas where you connect multiple AI models and tools into automated creative pipelines. Instead of switching between platforms or doing repetitive tasks manually, you build a graph of connected nodes—each performing one step—and run the whole sequence in one click.
## Why use Workflows?
Traditional AI design tools force you to choose between power and simplicity. You end up juggling multiple platforms or accepting limitations. Workflows removes these obstacles:
* **50+ models in one place.** Image, video, and text AI models are unified on a single canvas, eliminating the need to switch between tools.
* **Smart organization.** Use sections, pages, and folders to keep your workspace structured and clutter-free as your projects grow.
* **Collaboration.** Share your workflows with your team using a project link. Use built-in collaboration tools including shapes and sticky notes to communicate visually.
* **Node-based control.** A professional, node-based interface gives you precise control over every step of your pipeline—without requiring code.
## How it works
Every workflow is built from **nodes**—self-contained units that each perform one task. You connect nodes by dragging from an output handle on one node to an input handle on another. Data flows left to right across the canvas:
* **Input nodes** provide data (text prompts, uploaded images, audio files) to the next step.
* **Output nodes** generate or transform that data (a generated image, an upscaled video, a rewritten prompt).
* **Connections** are type-safe: image outputs connect to image inputs, text to text, video to video.
Once your nodes are wired together, click **Run** (or press `Ctrl/Cmd + Enter`) to execute the workflow. Each node processes in sequence, producing the final output from your input data.
Workflows are iterative by design. Adjust prompts, swap models, or rewire nodes between runs until the output matches your vision.
You can share any workflow with your team by sharing the project link from the Collaboration toolbar on the canvas.
## What you can build
* Generate an image from a prompt, upscale it to 4K, then resize it for every ad format in one run.
* Animate a still image into a video, extend it to a longer clip, add a voiceover with Lipsync, and export a finished asset.
* Feed a script through AI Copilot to expand it, split it into scenes, generate a video per scene, and stitch everything together.
* Batch-process dozens of product photos through the same editing prompt using Image Iterator.
## Explore the docs
Learn how to navigate the canvas, add and connect nodes, and organize your workspace.
Reference guide for every node type—image, video, text, and utility nodes.
Browse all 50+ image, video, and text models available on the canvas.
Open Workflows and start building.