# Generate images

Generate images in an image node from a text description or reference images, and adjust the image generation settings.

In an image node, you can **generate images from a text description**, or **use images from the Library as references**, to create the visuals you need.

After an image is generated, you can keep editing, adjusting, and refining it, then use the result in what you make next.

Create an image node yourself, or ask the Agent to create an image generation card.

![A new, empty image node with its prompt box below it](https://docs.moodio.art/media/8d105de8c19ed4359145f12793bb21d7.png)

## Generate images from text

In the image generation node, type what you want to generate. Use the prompt to describe the characters, scene, action, style, and other details of the picture, and the model generates an image from your description.

For example:

> A little girl sitting on a wooden pier by the sea, with a whale rising out of the water in the distance, in a warm and gentle picture-book style.

The more clearly you describe the subject, action, scene, and style, the more likely you are to get the result you expect.

![A character sheet generated from a text prompt, with the prompt in the prompt box below](https://docs.moodio.art/media/9b659306c90d24e64516660218b0f25a.png)

## Generate from reference images

Besides a text description, you can add **reference images or asset cards** that are already on the canvas. Type @ in the prompt box, or draw a connection from the media you want to reference to the generation node, and the model uses the visual information in the images when it generates.

![A reference image wired into a generation node and shown as a reference in its prompt box](https://docs.moodio.art/media/870f8ff1942da9756f964ff5e601bb44.png)

Reference images can provide:

- Character appearance
- Scene and setting
- Clothing and styling
- Color and composition
- Overall visual style

You can also find a suitable image in the **Library** and drag it onto the canvas to use as a reference. When you drag an image from the Library onto the canvas, a generation node that already references the image appears on the canvas automatically. Finish the prompt, and you can generate right away.

![An image dragged onto the canvas from the Discover panel, already referenced by a new generation node](https://docs.moodio.art/media/c64b03ba3606f5efebdd940092a997ef.png)

## Image generation settings

In the image generation node, you can adjust the generation settings to fit what you need.

[Opening the Image Size and Quality menus in an image node's prompt box and choosing settings](https://docs.moodio.art/media/1a2d75714c65d3d6d1c0f7f878f690aa.mp4)

- **Model**

Choose the AI model that generates the image. Models can differ in visual style, in the results they produce, and in other ways.

- **Aspect Ratio**

Set the aspect ratio of the image, such as landscape, portrait, or square. Choose a ratio that suits the video you plan to make or how you will use the image.

- **Image Size**

Set the size of the generated image. The size affects the resolution and dimensions of the final image.

- **Amount**

Set how many images to generate at a time. When you want several different results to compare, increase the number.

- **Quality**

In the image generation card, you can choose the generation quality that fits what you need: Smart, High, Medium, or Low.