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

# Generate images

> Create images from a prompt, refine them, and manage the results.

Open **Images → Generate image** in the sidebar. You can also ask for an image in a chat.

## Generate

1. Choose the **Image model**.
2. Write the **Prompt**: the subject, setting, colors, lighting, composition and mood. **Tips for better results** has examples.
3. Set the **Width** and **Height**. The note under the model says how it uses them: some models take the exact size, within a range; others use the aspect ratio at their own resolution, so pick a standard ratio such as 1024 × 1024 (square) or 1024 × 576 (16:9).
4. Pick a **Style**, or keep **No style**. The styles are your organization's.
5. Choose the **Output Format**. The formats depend on the model.
6. Choose **Generate Image**.

## Enhance your prompt

Write a rough idea in **Prompt**, then choose **Enhance prompt**. The assistant uses your **Image model**, **Style**, **Width** and **Height**.

If an important detail is unclear, it asks one question inside the prompt field. Choose an answer chip or type your answer, then choose **Enhance prompt** again. You can also press ⌘/Ctrl+Enter. After at most three questions, it writes a detailed prompt you can edit.

* **Skip questions:** write the prompt now, using sensible assumptions.
* **Start over:** restore your original draft and end the session.
* **Undo:** restore the text you had before enhancing, after the final prompt appears.

Any assumptions appear under the final prompt. Editing that prompt clears **Undo** and its assumptions. If enhancement fails, your text stays in the field so you can try again.

## Refine the result

Your image appears under **Generated Images** and is saved to the [Image library](https://docs.weelkit.com/images/image-library.md).

* To change it, describe what to change under **Refine Prompt** and choose **Regenerate Image**. Weelkit uses your current image as the starting point. Earlier images stay in the library.
* The image's **⋯** menu has:
  * **Regenerate:** make a new image from the prompt.
  * **Edit image:** open it in one of the [image tools](https://docs.weelkit.com/images/tools.md): **Crop**, **Resize**, **Change aspect ratio**, **Add overlay** or **Add watermark**. It shows the tools your organization has turned on, for roles that can save edited images.
  * **Delete:** asks you to confirm, then deletes the image and its stored files from the library for good.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.