How CanvasIQ works

Open the Workspace and start drawing — no account needed. The left panel holds tools and styling, the canvas sits in the middle, and the right panel has Properties, Layers and the AI Copilot. On phones the tools live in the bottom toolbar.

Your work stays in memory only (refreshing clears it). Use Save → Project to download a .canvasiq.json file and Open in the top bar to load it back anytime.

Undo / Redo: Ctrl/Cmd+Z and Ctrl+Y (or Ctrl/Cmd+Shift+Z) — up to 100 history states. Duplicate: Ctrl/Cmd+D. Select all: Ctrl/Cmd+A. Delete: Delete/Backspace.

The Layers tab lets you create, rename (double-click), hide, lock, reorder and delete layers. Group moves the current selection onto a fresh layer. Locked layers ignore all pointer input — handy for backgrounds and templates.

The Copilot (right panel → AI) has four modes:

• Draw — text to vector art: logos, silhouettes, floor plans, illustrations.
• Diagram — architecture, flowcharts and org charts laid out automatically.
• Shapes — smart geometry: wireframes, dashboards, device frames, product boxes.
• Assist — reviews your actual canvas and suggests layout, alignment, color and accessibility improvements.

Every AI response is structured JSON that is strictly validated and sanitized before any shape is created — raw AI output is never rendered. Generated artwork lands on its own layer, fully editable.

Switch the top-bar toggle to 3D. Add cubes, spheres, cylinders, cones and tori, or extrude an SVG path from your 2D canvas. Drag to orbit, scroll to zoom. Select an object to move/rotate/scale it with the gizmo, and tune color, metalness and roughness in Properties. Four lighting presets: Studio, Daylight, Sunset, Night.

Switch the top-bar toggle to Paint for a bitmap editor in the style of classic Paint. Strokes become pixels in the picture, so you change them with selections, undo and the image tools rather than by picking an object. Your Paint picture is kept when you switch to 2D or 3D and back. There is a fuller guide on the Paint page.

• Draw — pencil, nine brushes, eraser (right button swaps only Color 1 for Color 2), fill with tolerance, color picker, magnifier, text.
• Shapes — line, curve, polygon and 20 presets, each with outline and fill styles. Drag the handles to adjust before you place a shape.
• Select — rectangle or free-form. Drag to move, Ctrl or Option to drag a copy, Shift to leave a trail. Cut, copy, paste, crop, invert.
• Image — resize and skew, rotate, flip, invert, grayscale, black and white, sepia. Drag the edges of the page, or use File → Image properties, to change the canvas size.
• Colors — click a swatch for Color 1, right-click for Color 2, X swaps them. Edit colors adds up to ten custom colors.
• View — zoom 12.5% to 800%, a pixel grid from 400%, rulers and a status bar. Hold Space and drag to pan.

Save offers PNG, JPEG, WebP, BMP, GIF, TIFF and PDF; Open loads PNG, JPEG, GIF, BMP, WebP and SVG. You can also paste an image with Ctrl/Cmd+V or drop a file on the canvas. Closing the tab with unsaved changes asks first, because nothing is stored for you.

Save (top bar, or Ctrl/Cmd+S) downloads your design in any of these formats. Open (top bar, or Ctrl/Cmd+O) loads a previously saved Project file back into the workspace.

• PNG — 1–3× scale, optional transparent background.
• SVG — true vectors serialized from the shape model.
• PDF — A4 document with the design fitted and branded footer.
• Project (JSON) — the full document; the format Open understands.
• GLTF / OBJ / STL — your 3D objects, ready for engines and 3D printing.

All files are generated client-side. Nothing is uploaded anywhere.

Is my work saved? Only in memory while the tab is open — by design. Export a Project file to keep it.

Does it work in Incognito? Yes — CanvasIQ uses no cookies, no local storage and no tracking.

What does the AI see? Only the prompt you type (plus a structural summary of your canvas in Assist mode). Requests go through the CanvasIQ server so the API key is never exposed to your browser.