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

# Models

> Available AI models for video generation

## Overview

RawUGC offers 5 AI video generation models. Each model has different capabilities, credit costs, and required parameters.

| Model                      | Type                | Credits | Estimated Time |
| -------------------------- | ------------------- | ------- | -------------- |
| `sora-2-text-to-video`     | Text-to-Video       | 6       | 1-3 minutes    |
| `sora-2-image-to-video`    | Image-to-Video      | 6       | 2-4 minutes    |
| `kling-2.6/motion-control` | Motion Transfer     | 18      | 2-5 minutes    |
| `veo3`                     | Text/Image-to-Video | 6       | 2-4 minutes    |
| `veo3_fast`                | Text/Image-to-Video | 6       | 1-3 minutes    |

***

## Sora 2 Text-to-Video

Create videos directly from text prompts.

**Required parameters:** `prompt`

**Optional parameters:** `aspectRatio` (`portrait` | `landscape`), `nFrames` (`10` | `15`), `selectedCharacter`

```json theme={null}
{
  "model": "sora-2-text-to-video",
  "prompt": "A serene beach at sunset with gentle waves",
  "aspectRatio": "landscape",
  "nFrames": "15"
}
```

## Sora 2 Image-to-Video

Generate videos from source images with text guidance.

**Required parameters:** `prompt`, `imageUrls`

**Optional parameters:** `aspectRatio` (`portrait` | `landscape`), `nFrames` (`10` | `15`), `selectedCharacter`

```json theme={null}
{
  "model": "sora-2-image-to-video",
  "prompt": "Camera slowly zooms in on the product",
  "imageUrls": ["https://example.com/product.jpg"],
  "aspectRatio": "portrait",
  "nFrames": "10"
}
```

## Kling 2.6 Motion Control

Transfer motion from a reference video to a character image. Ideal for dance videos and character animation.

**Required parameters:** `imageUrls`, `videoUrls`, `characterOrientation`, `mode`

**Optional parameters:** `prompt` (max 2500 chars)

| Parameter              | Values           | Description                                                                                                   |
| ---------------------- | ---------------- | ------------------------------------------------------------------------------------------------------------- |
| `characterOrientation` | `image`, `video` | `image` = same orientation as person in image (10s max). `video` = consistent with video character (30s max). |
| `mode`                 | `720p`, `1080p`  | Output resolution                                                                                             |

```json theme={null}
{
  "model": "kling-2.6/motion-control",
  "imageUrls": ["https://example.com/character.jpg"],
  "videoUrls": ["https://example.com/dance-reference.mp4"],
  "characterOrientation": "video",
  "mode": "1080p"
}
```

<Note>
  Kling Motion Control costs 18 credits per video (3x other models) due to the complexity of motion transfer.
</Note>

## Veo 3.1 Quality

Google's flagship video model with highest fidelity. Supports both text-to-video and optional image-to-video.

**Required parameters:** `prompt`

**Optional parameters:** `imageUrls` (up to 2), `aspectRatio` (`16:9` | `9:16` | `Auto`)

```json theme={null}
{
  "model": "veo3",
  "prompt": "A cinematic drone shot of a mountain landscape at golden hour",
  "aspectRatio": "16:9"
}
```

<Tip>
  With 1 image, the video unfolds from it. With 2 images, they serve as the first and last frame.
</Tip>

## Veo 3.1 Fast

Cost-efficient variant of Veo 3.1 with faster generation and strong results. Same parameters as Veo 3.1 Quality.

**Required parameters:** `prompt`

**Optional parameters:** `imageUrls` (up to 2), `aspectRatio` (`16:9` | `9:16` | `Auto`)

```json theme={null}
{
  "model": "veo3_fast",
  "prompt": "A product showcase with smooth camera movement",
  "aspectRatio": "9:16"
}
```