> For the complete documentation index, see [llms.txt](https://docs.maiagent.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.maiagent.ai/maiagent-user-guide/en/work/maigpt/image.md).

# AI Image Generation

MaiGPT AI image generation: describe a scene in plain language to generate images, edit and iterate repeatedly, and tips for writing good prompts.

MaiGPT has built-in AI image generation. Describe the image you want in words, and the AI creates it for you. Descriptions in Chinese are supported, so you don't need to write in English. After an image is generated, you can keep refining it until you are satisfied.

***

## Generate Images from a Text Description <a href="#generate-from-text" id="generate-from-text"></a>

Describe the image you want directly in the conversation:

> "Draw an illustration of a modern office with bright colors"
>
> "Generate a marketing product image with a blue gradient background"
>
> "Draw a cute mascot: a cat in a suit"

### Prompt Tips <a href="#prompt-tips" id="prompt-tips"></a>

The more specific your description, the closer the image will be to what you expect:

| Generic description | Better description                                                                                                  |
| ------------------- | ------------------------------------------------------------------------------------------------------------------- |
| "Draw a cat"        | "An orange cat sitting on a windowsill, sunset in the background, watercolor style"                                 |
| "Make a poster"     | "A vertical A4 poster titled 'Annual Conference 2025', tech style, blue-purple tones"                               |
| "Product image"     | "A cup of coffee on a white background, top-down angle, a few coffee beans beside it, commercial photography style" |

You can specify:

* **Style** — watercolor, flat design, realistic, cartoon, commercial photography
* **Color tone** — bright, warm, cool, monochrome
* **Composition** — top-down, front view, close-up, wide shot
* **Purpose** — social media post, presentation illustration, marketing material

### Download an Image <a href="#download-image" id="download-image"></a>

After an image is generated, click it to view it enlarged. To save it to your device:

* **Computer**: while viewing the image, right-click
* **Phone / tablet**: while viewing the image, long-press it

***

## Edit and Iterate on Images <a href="#edit-and-iterate" id="edit-and-iterate"></a>

Not happy with the generated image? You don't need to start over. You can keep revising it in the same conversation until you are satisfied.

### How to Edit <a href="#how-to-edit" id="how-to-edit"></a>

After an image is generated, simply tell the AI what to change in the conversation:

> "Change the background to white"
>
> "Make the person's expression a smile"
>
> "Add the company logo in the bottom-right corner"
>
> "Make the overall color tone a bit warmer"

The AI generates a new version based on your instructions.

### Generate from an Existing Image <a href="#generate-from-existing-image" id="generate-from-existing-image"></a>

You can also upload an image and have the AI use it as a reference to generate a new one:

> (Upload a product photo)
>
> "Use the style of this image as a reference and generate similar marketing material"
>
> "Turn this image into a cartoon style"

### Tips <a href="#tips" id="tips"></a>

{% hint style="success" %}
**Edit step by step**

Don't ask for too many changes at once. Changing one thing at a time works best:

1. First make sure the composition and subject are right
2. Then adjust the color tone and style
3. Finally add details (text, logo, etc.)
   {% endhint %}

{% hint style="info" %}
**Start over if you're not satisfied**

If the edits keep drifting in the wrong direction, it's better to describe the image again from scratch. The AI generates independently each time, so starting over is sometimes faster than continuing to revise.
{% endhint %}

***

## Next Steps <a href="#next-steps" id="next-steps"></a>

* [Connectors](/maiagent-user-guide/en/work/maigpt/connectors.md)
* [Back to MaiGPT Overview](/maiagent-user-guide/en/work/maigpt.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.maiagent.ai/maiagent-user-guide/en/work/maigpt/image.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.