Skip to main content
AI Canvas is an infinite visual workspace. Every image and video you generate in Runable lives here. You can upload your own files, generate new ones from text, arrange them freely, and use precision editing tools directly on the canvas, all without sending prompts to the agent. The canvas is the shared workspace behind AI Images, AI Videos, and AI Carousel. Anything you create with those features appears on the canvas automatically.
AI Canvas showing multiple images and video clips arranged on a freeform workspace with the toolbar visible at the bottom

​
What you can do on the canvas

Upload and arrange

Drag and drop images and videos onto the canvas. Move, resize, and organize freely.

Generate images

Create images from a text prompt directly on the canvas with any model.

Generate videos

Create videos from text or from an image directly on the canvas.

Edit with precision

Remove backgrounds, upscale, edit text, and mark regions for targeted changes.

Combine into timeline

Stitch images and video clips into a single video output.

Download and export

Download individual files or batch export as a ZIP archive.

​
Canvas editing tools

When you select an image or video on the canvas, a toolbar appears with direct editing operations. These work without the agent and use significantly fewer credits.

​
Image tools

  • Remove BG - Strip the background
  • Text Edit - Detect and change text
  • Mark Edit - Draw on regions and describe changes
  • Upscale - Increase resolution 1x, 2x, 3x, or 4x
  • Edit Image - Prompt-based editing with any model
  • Create Video - Animate the image into a video

​
Video tools

  • Edit Video - Change style or content with a prompt
  • Motion Control - Transfer motion from a reference video
  • Remove BG - Strip the video background frame by frame
Using canvas editing tools directly gives you more control and uses significantly fewer credits than sending prompts to the agent. Use canvas edits for fine-tuning and save agent requests for generating new content or complex multi-step work.

​
Non-destructive editing

Every edit operation creates a new item on the canvas. Your original is always preserved. This gives you an implicit version history of every change you have made.

​
Upload limits

LimitValue
Maximum files per upload20
Maximum file size25 MB per file
Supported typesAll image and video formats

​
What AI Canvas does not support

  • Layers or layer ordering within the canvas. Items are arranged spatially, not stacked.
  • Drawing freehand on the canvas (pen, brush, shape tools). Use Mark Edit for region-based changes.
  • Adding text annotations directly on the canvas.
  • Real-time collaboration with other users on the same canvas.

​
Next steps

Canvas Basics

Learn how to upload, arrange, zoom, and navigate the canvas.

Generate Images on Canvas

Create images from text directly on the workspace.