> ## Documentation Index
> Fetch the complete documentation index at: https://docs.holala.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Ghost Mannequin

> Remove the mannequin from a garment photo while keeping its hollow, 3D shape and drape.

Ghost Mannequin removes the mannequin from your garment photo while keeping its hollow, three-dimensional shape and drape.

## What it does

This template takes a photo of a garment shot on a mannequin and produces a clean ghost-mannequin image — the mannequin disappears, but the garment keeps its dimensional form, collar structure, and natural drape, just as if it were photographed with the traditional invisible-mannequin technique.

## Inputs

* **Product image** (required) — a photo of the garment on a mannequin, with the shape and drape clearly visible. See [Product Images](/input-uploads) for prep guidelines.

Ghost Mannequin doesn't use a reference image or model selection — it works from your garment shot alone.

<Note>
  The AI refines what your photo shows — it can't invent structure the input doesn't reveal. If the mannequin photo is shot from the front, the result will be a front view.
</Note>

## How to use it

<Steps>
  <Step title="Open Ghost Mannequin in Studio">
    From [Studio](/studio), select the Ghost Mannequin template.
  </Step>

  <Step title="Upload your garment photo">
    Add the mannequin shot you want converted.
  </Step>

  <Step title="Set resolution and variations">
    Choose 1K for a quick test, or 2K/4K for final output. Pick how many variations to generate — more variations and higher resolution use more credits. See [Credits](/credits).
  </Step>

  <Step title="Generate">
    Results save automatically to your Library, ready to download as JPG, PNG, or WebP.
  </Step>
</Steps>

## Tips

<AccordionGroup>
  <Accordion title="Make sure the collar and neckline are fully visible">
    These areas define the garment's structure in the final result. If they're obscured or cropped in the source photo, the AI has nothing to work from.
  </Accordion>

  <Accordion title="Use a plain, contrasting background">
    A simple background helps the AI cleanly separate the garment from the mannequin, producing a crisper hollow-shape result.
  </Accordion>

  <Accordion title="Shoot the garment fully buttoned or zipped">
    A closed garment holds its shape better on the mannequin, which carries through to a more natural drape in the output.
  </Accordion>
</AccordionGroup>

<Frame caption="Before: garment shot on a visible mannequin. After: the same garment with the mannequin removed and its 3D shape retained.">
  ![Garment on a mannequin converted into a ghost-mannequin image](https://newpr.holala.ai/api/media/file/holala-fashion-ai-image-generation-tool-16.png)
</Frame>

## Related

<CardGroup cols={2}>
  <Card title="Product Images" icon="image" href="/input-uploads">
    Get your product photos ready before generating.
  </Card>

  <Card title="Model Remover" icon="user-slash" href="/templates/model-remover">
    Have an on-model shot instead of a mannequin? Remove the model and keep the product clean.
  </Card>
</CardGroup>