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

# Image to Prompt

> Reverse-engineer an image into a detailed generation prompt.

**Image to Prompt** (internally, *Promptify*) reverse-engineers an image into
a detailed text prompt you can feed to an image generator or LLM to recreate
it.

## Open it

Press <kbd>⌃</kbd> <kbd>⌥</kbd> <kbd>P</kbd> from any app.

## How it works

<Steps>
  <Step title="Copy an image">
    Copy any image — a design, a screenshot, something you want to reproduce.
  </Step>

  <Step title="Press the shortcut">
    Press <kbd>⌃</kbd> <kbd>⌥</kbd> <kbd>P</kbd>. The image is sent to a
    **vision model** with your chosen prompt style.
  </Step>

  <Step title="Paste the prompt">
    A detailed recreation prompt **lands on your clipboard** and appears in a
    panel. Paste it into any image generator or LLM.
  </Step>
</Steps>

## Prompt styles

Settings → Tools → Image to Prompt:

* **UI Focused** — a prompt to rebuild the screen in code.
* **General Purpose** — a balanced, faithful description.
* **Image Focused** — geared toward recreating the image in an image model.
* **Ask me every time** — pick the style from a menu on each run.

<Tip>
  Pair it with [Generate Image](/tools/generate-image): describe an image you
  like, then riff on that prompt to make a variation.
</Tip>

## Why a keypress, not a watcher

Image to Prompt is deliberately hotkey-triggered rather than watching your
clipboard, so images only ever leave the machine on an explicit press.