Generation tools

These tools live under Core Generation Tools in the canvas sidebar. Each opens a panel on the left; fill it in and press Send. New to the canvas? Read Using the canvas first.

Concept CreatorCtrl/⌘ + 1

Turn a description — or a technical sketch — into a finished, photorealistic garment. Cost: about 2 credits.

1. Open Concept Creator and keep Photorealistic selected under Select Workflow.
2. Add a reference image (optional): click Library, pick a technical sketch or product photo, and Insert it. It appears as your Reference image.
3. Type a Prompt describing the final look — e.g. Render this sketch as a photorealistic camel wool overcoat, editorial studio shot. Optionally set Ratio and Size.
4. Press Send. The rendered garment is placed on your canvas.

Use Photorealistic mode with a sketch as the reference to convert a flat drawing into a finished product shot.

Mix & MatchCtrl/⌘ + 2

Dress a model in a set of garments to preview a complete styled outfit. Cost: about 1 credit.

1. Open Mix & Match.
2. Click Library and add your garments top to bottom, outer to inner — e.g. select the top first, then the bottom — and Insert. Use images over 1080px for best results.
3. Adjust the pre-filled Prompt if you want a specific model or setting (it defaults to a fashion-editorial scene).
4. Press Send to generate the styled look on a model.

Style TrainingCtrl/⌘ + 4

Train a custom style (a Flux/LoRA model) from a set of reference images so future generations match your house look. Cost: about 100 credits.

1. Open Style Training.
2. Add a set of images that share a consistent style via Browse or Library (more, varied angles train better).
3. Enter a Style Name you'll recognise later.
4. Press Send to start training.

Training is asynchronous and takes several minutes. Credits are charged up front; you'll be notified when the style is ready to use.

Video GenerationCtrl/⌘ + 5

Animate a still into a short fashion clip — a runway walk, a turn, or a fabric-in-motion shot. Cost: about 40 credits.

1. Open Video Generation and add a start image via Browse or Library (optional but recommended).
2. Give the clip a Video Name and a Prompt describing the motion — e.g. The model walks toward the camera down a softly lit runway. Tip: use AI Generate Video Idea for a starting prompt.
3. Choose an Aspect Ratio and tweak Advanced Settings if needed.
4. Press Send to start rendering.

Video rendering is asynchronous and takes a few minutes. The clip appears on your canvas when it's ready.

UpscaleCtrl/⌘ + 6

Increase an image's resolution — up to 8 megapixels — for print and high-detail presentation. Cost: about 1 credit.

1. Open Upscale and add the image via Library, Browse, or by dragging it from the canvas.
2. Choose a Target Resolution (1–8 MP) and, optionally, enable Enhancements.
3. Press Send. The upscaled version is placed on your canvas.

Remove BackgroundCtrl/⌘ + 7

Cut the subject out of its background to get a clean, transparent cut-out. Cost: about 1 credit.

1. Open Remove Background.
2. Add the image via Library, Browse, or drag it from the canvas.
3. Press Send to get the background-free cut-out.

Image to SketchCtrl/⌘ + 8

Convert a garment photo into a clean technical flat sketch through a three-stage pipeline. Cost: about 1 credit.

1. Open Image to Sketch and add a garment photo via Library or Browse.
2. Review the three pipeline stages — Flat Image, CAD Image, and Technical Sketch.
3. Press Start Pipeline (or Send) to produce the technical flat.

Split Layer (Beta)

Separate an image into stacked visual layers — for example to isolate the subject from the background or break a look into pieces. Cost: about 2 credits.

1. Open Split Layer.
2. Pick the Number of Layers (2–10) to separate the image into.
3. Add the image via Library or Browse.
4. Press Send to get the separated layers on your canvas.

Split Layer is in beta; results improve with clean, well-lit source images.