Canvas

The canvas is where your creative work lives. It’s an infinite, zoomable workspace where you place reference images, arrange generated content, and build visual narratives spatially — instead of scrolling through a linear chat history.

​

Why a canvas

Most AI generation tools work like a chat window: you type a prompt, get a result, type another, and scroll back to find something from earlier. The canvas changes this by giving you a spatial workspace where every element stays visible, movable, and organized the way you want. This matters because creative work is nonlinear. You might want to compare five variations of a scene side by side, group character references together, or arrange a storyboard sequence across the canvas. The spatial layout makes these workflows natural.

​

Canvas tools

The toolbar at the top of the canvas provides:

• Select — Click and drag to select, move, and resize items
• Pan — Hold and drag to move around the canvas, or scroll to zoom
• Assets — Open the asset panel to upload and manage your reference material
• Upload — Quickly upload an image or video file to the canvas
• Text — Add text labels and annotations
• Note — Add sticky notes for creative direction or reminders
• Shapes — Access geometric shapes for visual planning

​

Shapes

The canvas includes geometric shapes for visual planning and layout: Rectangles, ellipses, triangles, diamonds, stars, hexagons, clouds, hearts, and more. Use these to sketch layouts, mark areas of interest, or create visual guides for your compositions.

​

Working with assets

Click the Assets button in the toolbar to open the asset panel. From here you can:

• Browse your workspace asset library with tag and EXIF filters
• Upload images, videos, and audio files from your device
• Drag assets onto the canvas as references
• Tag assets directly from the canvas toolbar

Uploaded assets are automatically clustered by visual similarity and tagged by AI. You can also add your own tags for custom organization. See Asset library for details.

​

Using skills on the canvas

Type / in a generation prompt field to invoke a skill — a reusable prompt template. The skill’s content is appended to your prompt, so you can apply consistent style directions or composition rules without retyping them.

​

Keyboard shortcuts

The canvas supports standard shortcuts for fast navigation and editing: Scroll to zoom, and hold Space to pan. For a complete list, open the Menu on the canvas and select Keyboard shortcuts.

​

Selection and interaction

• Click an item to select it
• Shift + Click to add or remove items from a selection
• Drag on empty space to create a selection box
• Cmd + Enter with image or video shapes selected to focus the chat panel with those items as context

​

Frames

Frames are container shapes that group content visually. Drag items into a frame to organize them — when you move the frame, everything inside moves together. Use frames to group:

• Character reference sheets
• Scene variations for comparison
• Storyboard sequences

To create a frame, open the Shapes menu in the toolbar and select Frame.

​

Spatial organization tips

• Group related content — Place character references together, scene images in another area, style references in another
• Use frames — Frames act as containers that group content visually
• Zoom levels — Zoom out for the big picture, zoom in for detail work
• Notes for context — Add sticky notes near images to capture creative direction that the AI agent can reference