# Generate videos

Choose a generation model and a video generation mode in a video node, and adjust generation settings such as duration, aspect ratio, resolution, draft, and audio.

In a video node, you can generate videos from **a text description, images, or video footage**, and choose the video generation mode that fits what you want to make.

After a video is generated, you can capture frames from it, trim it, upscale it with Super resolution, separate its audio and video, and more, then use the results in what you make next.

Create a video node yourself, or ask the Agent to create a video generation card.

![A new, empty video node; its prompt box shows the model, the generation mode, and the settings](https://docs.moodio.art/media/3933fa7706f4bfacd989a3adf72dcb6a.png)

## Choose a generation model

In the video generation card, click the model name to see and choose from the video generation models. Models can differ in the results they produce and in the features and settings they support, so choose the one that fits what you want to make.

## Choose a **Generation mode**

In the video generation node, you can choose the video generation mode that fits your needs. Each mode suits a different kind of video.

### **Text to video**

Generates a video directly from a text description.

Describe the characters, scene, action, camera shots, and so on, and the AI generates a video from your prompt.

Use it when you have no images or video to start from and want to go straight from an idea to a video.

### **Omni reference**

Uses images and other references to guide the AI as it generates the video.

References can show what characters, scenes, objects, and so on look like, while the prompt controls the action and the picture in the video.

Use it when the video needs to keep a specific character, scene, or visual element consistent.

### **Video editing**

Starts from an existing video and changes or adjusts its content from a text description.

You can change part of the video while keeping as much of the rest of the original as possible.

Use it when you already have a video but want to change something in it.

### **Video extension**

Starts from an existing video and generates more of it, making the original video longer.

For example, after a video ends, you can generate the next action or scene so that the video continues naturally.

Use it when you want a longer video, or want to show what happens after the original video ends.

### **First & last frame**

Set the video's **first frame and last frame** to generate a video of the change from one image to the other.

You choose the picture at the start of the video and the picture at the end, and the AI generates the motion in between from the two images.

Use it when you already know how the video starts and ends and want to generate the transition between them.

## Adjust the video generation settings

Adjust the video generation settings to fit what you need.

[In a video node's prompt box, checking the model, switching the generation mode, and setting duration, aspect ratio, resolution, and audio](https://docs.moodio.art/media/1d595c2741fead0e2e291cd2ff27da37.mp4)

- **Duration (seconds)**

Set how long the generated video is. Choose a length that suits the content and how you plan to edit it later. The shortest and longest lengths can differ between models. You can also turn on Smart to let the model decide the length.

- **Aspect Ratio**

Set the aspect ratio of the video, such as landscape, portrait, or square. Choose the ratio that fits where the video will be used in the end.

- **Resolution**

Set the output resolution of the generated video. The resolution affects how sharp the video is and the resources needed to generate it.

- **Draft**

Use draft mode when you want to compare motion, composition, or prompts before you decide whether to generate the final video. A 480p draft costs less, so you can make one first to quickly check the motion, the picture, and the overall result. Testing in draft mode first keeps you from wasting credits on generating final videos again and again.

When the draft is done, the video node shows a **Draft** tag and offers **Generate final**, which reuses the draft's prompt and media. Choose the resolution of the final video (1080p by default), check the estimated credits, and generate.

[After a draft video is done, Generate final starts the final video on a new node](https://docs.moodio.art/media/98576d5f2bb594e16d5900d62abbc2cc.mp4)

Generating the final video creates a new final video node on the canvas, and the original draft is kept.

- **Generate Audio**

Choose whether to generate audio along with the video. When it is on, the generated video can include matching audio. When it is off, only the picture is generated.