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

# AI Videos

> Generate videos from text or images, edit existing videos, remove backgrounds, control motion, and combine clips into a timeline.

AI Videos lets you generate video clips from text prompts or still images, edit existing videos with natural language, remove video backgrounds, transfer motion between sources, and combine multiple clips into a single timeline. All videos live on the [AI Canvas](/overview-what-is-ai-canvas) alongside your images and other media.

<Frame caption="Generated video clips displayed on the canvas workspace.">
  ![AI Canvas showing several generated video clips with thumbnails and play buttons visible](https://runable.cloud/assets/video-q9i8rVvo.png)
</Frame>

## What you can do

<CardGroup cols={3}>
  <Card title="Generate from text" icon="video" href="/generate-from-text">
    Describe a scene and the AI creates a video clip with motion and audio.
  </Card>

  <Card title="Generate from an image" icon="image" href="/generate-from-an-image">
    Turn a still image into an animated video clip.
  </Card>

  <Card title="Edit a video" icon="pen-to-square" href="/edit-a-video">
    Change the style, color grade, or content of a video with a prompt.
  </Card>

  <Card title="Remove background" icon="scissors" href="/remove-video-background">
    Strip the background from a video, leaving the subject isolated.
  </Card>

  <Card title="Motion control" icon="person-walking" href="/motion-control">
    Transfer motion from a reference video onto a still image.
  </Card>

  <Card title="Combine into timeline" icon="film" href="/combine-into-timeline">
    Arrange multiple clips and images into a single video output.
  </Card>
</CardGroup>

Additional capabilities:

<AccordionGroup>
  <Accordion title="A→B transitions - Start and end frame control">
    Provide both a starting image and an ending image. The AI generates a smooth transition between the two states. Use this for product reveals, before-and-after sequences, or any scene that moves from one defined state to another.
  </Accordion>

  <Accordion title="Element injection - Add characters or objects to a video">
    Upload a frontal image of a character or object. Reference it in your prompt with @Element1 syntax. The AI injects it into the video while keeping the original motion and scene.
  </Accordion>

  <Accordion title="Style reference - Guide the visual look with an image">
    Attach a style reference image. The AI matches the color palette, lighting, and visual treatment to create a consistent look across your video.
  </Accordion>

  <Accordion title="Audio generation - Native lip-sync and sound">
    Enable audio during generation for dialogue, singing, or ambient sound. The AI creates synchronized audio that matches the visual content. Enabling audio costs approximately 2x the credits of a muted video.
  </Accordion>

  <Accordion title="Long video - Multi-shot sequences stitched together">
    For videos longer than 15 seconds, the agent plans individual shots, generates each clip, and stitches them together using the timeline. This requires your approval before generation since it uses more credits.
  </Accordion>
</AccordionGroup>

## Options you can configure

Before generating, you can set:

| Option           | Choices                                         |
| ---------------- | ----------------------------------------------- |
| **Aspect ratio** | 21:9, 16:9 (default), 4:3, 1:1, 3:4, 9:16, 9:21 |
| **Duration**     | 3–15 seconds (varies by model)                  |
| **Resolution**   | 480p, 720p, 1080p, 4K (varies by model)         |
| **Audio**        | On or off. Muted videos use half the credits.   |

<Frame caption="The video metadata options showing aspect ratio selector and audio toggle.">
  ![Home input showing the aspect ratio dropdown set to 16:9 and an audio toggle button with a tooltip explaining muted videos use half the tokens](https://runable.cloud/assets/1-Cu0ZTQTL.png)
</Frame>

## Video generation models

Multiple AI models are available across 7 families. The agent selects the best model for your task automatically.

| Family             | Strengths                                           | Text-to-video | Image-to-video | Editing     | Audio |
| ------------------ | --------------------------------------------------- | ------------- | -------------- | ----------- | ----- |
| **Kling 2.6/3.0**  | Most versatile, A→B transitions, element injection  | Yes           | Yes            | Yes (O1/O3) | Yes   |
| **Veo 3.1**        | High quality, up to 4K resolution                   | Yes           | Yes            | No          | Yes   |
| **Seedance 1.5**   | Widest aspect ratio support, available on Free plan | Yes           | Yes            | No          | Yes   |
| **MiniMax Hailuo** | Fast generation                                     | Yes           | Yes            | No          | No    |
| **Wan v2.6**       | Open-source, low cost                               | Yes           | Yes            | No          | No    |

## Non-destructive editing

Every video operation (edit, motion control, background removal) creates a new video on the canvas rather than overwriting the original. Your original video is always preserved.

## What AI Videos does not support

* Frame-by-frame editing or manual keyframing.
* Adding text overlays or subtitles directly. Use an external editor for post-production text.

## Next steps

## Next steps

<CardGroup cols={2}>
  <Card title="Generate from Text" icon="play" href="/generate-from-text">
    Create your first video from a text prompt.
  </Card>

  <Card title="Generate from an Image" icon="image" href="/generate-from-an-image">
    Turn a still image into an animated video.
  </Card>
</CardGroup>