> 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/maiagent-user-guide-en/tools/gemini-3-image-tool.md).

# Gemini 3 Image Generation Tool

### Feature Overview <a href="#feature-introduction" id="feature-introduction"></a>

The Gemini 3 Image Generation Tool is an AI image generation feature powered by MaiAgent's integration with the Google Gemini 3 Pro Image model. This tool enables AI Assistants to generate high-quality images based on users' text descriptions, supports both Traditional Chinese and English prompts, and allows style transfer and image editing through reference images.

{% hint style="info" %}
Gemini 3 Pro Image is Google's latest multimodal AI model with powerful image generation capabilities. It can generate high-quality, contextually appropriate images based on contextual understanding and world knowledge reasoning.
{% endhint %}

***

### Key Features <a href="#key-features" id="key-features"></a>

<table><thead><tr><th width="200">Feature Type</th><th>Description</th><th>Use Cases</th></tr></thead><tbody><tr><td><strong>Text-to-Image</strong></td><td>Generate images based on text descriptions</td><td>Marketing materials, social media posts, product visualization</td></tr><tr><td><strong>Reference Image Editing</strong></td><td>Perform style transfer or content modification based on reference images</td><td>Brand visual consistency, image style adjustment</td></tr><tr><td><strong>Conversational Image Editing</strong></td><td>Continuously adjust and refine images across multi-turn conversations</td><td>Iterative design, detail fine-tuning</td></tr><tr><td><strong>Chinese Prompt Support</strong></td><td>Directly describe desired images in Traditional Chinese</td><td>Chinese-speaking users can use it without translation</td></tr></tbody></table>

***

### Enable the Gemini 3 Image Generation Tool <a href="#enable-gemini-3-image-tool" id="enable-gemini-3-image-tool"></a>

#### Prerequisites <a href="#prerequisites" id="prerequisites"></a>

1. Confirm your organization plan supports image generation
2. Confirm AI feature permissions are enabled
3. Prepare the AI Assistant that will use this tool

#### Enable Steps <a href="#enable-steps" id="enable-steps"></a>

1. **Go to the AI Assistant Settings Page**
   * Click "<mark style="color:blue;">AI Assistant</mark>" in the left menu
   * Select the AI Assistant to configure
   * Click "<mark style="color:blue;">Settings</mark>"
2. **Configure Tool Settings**
   * Switch to the "<mark style="color:blue;">Tools</mark>" tab
   * Click the "<mark style="color:blue;">Select Tools</mark>" button
   * Find "<mark style="color:blue;">Nano Banana Pro Image Generation</mark>" in the available tools list and check it
   * Click "<mark style="color:blue;">Confirm</mark>"
3. **Save Settings**
   * Click the "<mark style="color:blue;">Save</mark>" button in the bottom right corner
   * Wait for the system to confirm the settings are complete

{% hint style="warning" %}
**Notes**

* Image generation consumes token quota. Please be aware of your plan limits
* Generated images default to 4K high-quality output
  {% endhint %}

***

### How to Use <a href="#how-to-use" id="how-to-use"></a>

#### Use via Chat Interface <a href="#use-via-chat-interface" id="use-via-chat-interface"></a>

1. **Enter an Image Description**
   * Type your desired image description in the chat box
   * Examples:
     * "Generate a landscape image of a sunset beach for me"
     * "Draw a cat wearing a spacesuit"
     * "Design a minimalist-style coffee shop logo"
     * "Create a watercolor painting of a mountain landscape"
2. **Use Reference Images**
   * Upload an image in the conversation as a reference
   * Combine with text instructions to describe the desired modifications
   * Examples:
     * Upload a photo and say: "Convert this photo to a watercolor painting style"
     * Upload a product image and say: "Design a promotional poster for this product"
3. **Conversational Editing**
   * After generating an image, request modifications in subsequent messages
   * Examples:
     * "Change the background to blue"
     * "Add some stars"
     * "Make the overall atmosphere warmer"

{% hint style="info" %}
The more detailed the description, the better the generated image quality. It is recommended to include details such as style, color, composition, and atmosphere.
{% endhint %}

***

### Use Case Examples <a href="#use-case-examples" id="use-case-examples"></a>

#### Scenario 1: Marketing Material Creation <a href="#use-case-marketing-materials" id="use-case-marketing-materials"></a>

**Requirement**: Quickly generate visual materials for social media posts

**Conversation Example**:

```
User: Generate an Instagram-suitable image for me. The theme is spring new product launch, with a fresh and bright style, featuring cherry blossom and shopping bag elements.
AI: [Generated image] Created a spring new product launch promotional image for you, featuring cherry blossom petals and stylish shopping bag elements with an overall fresh and bright style.
User: Great, but can you change the shopping bag color to pink?
AI: [Generated updated image] The shopping bag has been adjusted to pink, making it more consistent with the cherry blossom theme.
```

#### Scenario 2: Product Visualization <a href="#use-case-product-visualization" id="use-case-product-visualization"></a>

**Requirement**: Generate visual representations for product concepts not yet in production

**Conversation Example**:

```
User: I want to design a smartwatch with a round watch face, silver metal band, and deep blue dial. Please generate a concept image for me.
AI: [Generated image] Created a smartwatch concept image with a round watch face paired with a silver metal band and deep blue dial.
```

#### Scenario 3: Style Transfer <a href="#use-case-style-transfer" id="use-case-style-transfer"></a>

**Requirement**: Convert existing images to different artistic styles

**Conversation Example**:

```
User: [Upload photo] Convert this photo to comic style
AI: [Generated image] Your photo has been converted to comic style, preserving the original composition and main elements.
```

***

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

#### Effective Prompt Writing <a href="#effective-prompt-writing" id="effective-prompt-writing"></a>

**Include specific descriptions**:

* Subject: What content you want to generate
* Style: Realistic, cartoon, watercolor, oil painting, graphic design, etc.
* Color: Main color tones and color preferences
* Composition: Close-up, panoramic, bird's-eye view, etc.
* Atmosphere: Warm, cool, lively, calm, etc.

**Good examples**:

* "An orange Shiba Inu sitting on a park bench in autumn, with golden maple leaves in the background, warm afternoon light, realistic photography style"
* "Minimalist-style tech company logo, using blue and white, geometric design"

**Avoid being too vague**:

* "Draw a picture" (lacks specific description)
* "A nice image" (no clear direction)

***

### FAQ <a href="#faq" id="faq"></a>

#### Q: What languages does the Gemini 3 Image Generation Tool support for prompts? <a href="#faq-supported-languages" id="faq-supported-languages"></a>

**A**: It supports Traditional Chinese and English prompts. You can directly describe the desired image content in Chinese.

#### Q: How many tokens does it cost to generate one image? <a href="#faq-token-cost-per-image" id="faq-token-cost-per-image"></a>

**A**: Image generation consumes more tokens than text-only conversations. The actual consumption depends on the prompt complexity and image quality settings. It is recommended to monitor your organization's token usage to avoid exceeding plan limits.

#### Q: Can I use reference images? <a href="#faq-reference-image" id="faq-reference-image"></a>

**A**: Yes. Upload an image in the conversation as a reference, and the AI Assistant will incorporate it into the generation process. This is especially useful for style transfer, image editing, or creating based on existing materials.

#### Q: What is the resolution of generated images? <a href="#faq-image-resolution" id="faq-image-resolution"></a>

**A**: Images are generated in 4K high quality by default, suitable for most use cases.

#### Q: Can I continuously modify images across multi-turn conversations? <a href="#faq-multi-turn-editing" id="faq-multi-turn-editing"></a>

**A**: Yes. The AI Assistant remembers the conversation context. You can request detail adjustments, style modifications, or element additions in subsequent messages to progressively refine the image.

#### Q: What should I do if image generation fails? <a href="#faq-generation-failure" id="faq-generation-failure"></a>

**A**: Possible causes and solutions:

* **Prompt violates content policy**: Adjust the description to avoid inappropriate content
* **System is busy**: Try again later
* **Insufficient token quota**: Check the remaining quota of your organization plan

***

### Related Documentation <a href="#related-docs" id="related-docs"></a>

{% hint style="info" %}
**Further Reading**

* [Tool Feature Overview](/maiagent-user-guide/maiagent-user-guide-en/tools/tool_description.md)
* [Configure Tools for AI Assistants](/maiagent-user-guide/maiagent-user-guide-en/tools/configure_tools.md)
* [Built-in AI Image Generation Tool](/maiagent-user-guide/maiagent-user-guide-en/tools/ai-image-generation-guide.md)
* [Create an AI Assistant](https://github.com/Playma-Co-Ltd/maiagent-user-guide-gitbook/tree/main/en/build/setup.md)
  {% endhint %}


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.maiagent.ai/maiagent-user-guide/maiagent-user-guide-en/tools/gemini-3-image-tool.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.