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

# Video Generation

> Generate videos from text descriptions using AI video models

## Overview

EnConvo brings AI video generation to your desktop. Create videos from text prompts or reference images using cutting-edge models like OpenAI Sora, Google Veo, Kling, Hailuo, and Wan -- all without leaving your workflow.

## Supported Models

| Model                          | Provider              |  Text-to-Video  | Image-to-Video | Strengths                                           |
| ------------------------------ | --------------------- | :-------------: | :------------: | --------------------------------------------------- |
| **Sora-2**                     | OpenAI                |       Yes       |       Yes      | Cinematic quality, realistic motion                 |
| **Sora-2 Pro**                 | OpenAI                |       Yes       |       Yes      | Higher quality, more detail                         |
| **Veo3**                       | Google (via fal.ai)   |       Yes       |       Yes      | Photorealistic, natural motion                      |
| **Veo3 Fast**                  | Google (via fal.ai)   |       Yes       |       Yes      | Faster generation, good quality                     |
| **Kling Video v2.5 Turbo Pro** | Kuaishou (via fal.ai) |       Yes       |       Yes      | Fast, high quality, good value                      |
| **Hailuo 02 Standard**         | MiniMax (via fal.ai)  |       Yes       |       Yes      | Versatile, affordable                               |
| **Hailuo 02 Pro**              | MiniMax (via fal.ai)  |       Yes       |       Yes      | Premium quality                                     |
| **Wan 25 Preview**             | Alibaba (via fal.ai)  |       Yes       |       Yes      | Chinese-developed, versatile                        |
| **xAI video features**         | xAI                   | Model-dependent |       Yes      | xAI media workflows when available for your account |

## Getting Started

<Steps>
  <Step title="Choose a Provider">
    Open the **Text to Video** or **Image to Video** command settings. Select your **Video Generation Provider**:

    * **Enconvo Cloud Plan** -- use multiple models with your Enconvo points
    * **OpenAI** -- use your own OpenAI API key for Sora models
    * **Fal.ai** -- use your own fal.ai API key for Kling, Hailuo, Veo, and Wan
  </Step>

  <Step title="Select a Model">
    Choose the model that fits your needs. Kling Video v2.5 Turbo Pro offers the best balance of speed, quality, and cost.
  </Step>

  <Step title="Configure Settings">
    Set resolution and duration options based on the selected model.
  </Step>

  <Step title="Generate">
    Enter a text prompt or provide a reference image, and let the AI create your video.
  </Step>
</Steps>

## Text to Video

Generate videos from text descriptions alone.

### How to Use

1. Open the **Text to Video** command from SmartBar or the command list
2. Enter a descriptive prompt in English
3. Wait for the video to generate (typically 30 seconds to a few minutes)
4. View, save, or share the resulting video

### Writing Effective Prompts

<Tip>
  Prompts must be in **English** for best results across all models.
</Tip>

**Basic structure:**

```
[Subject] + [Action] + [Setting] + [Style/Mood] + [Camera Movement]
```

**Example prompts:**

```
A golden retriever running through a field of wildflowers at sunset,
slow motion, cinematic lighting, warm tones
```

```
A futuristic city skyline at night with flying cars and neon lights,
drone shot sweeping across the buildings, cyberpunk aesthetic
```

```
A cup of coffee being poured in slow motion, close-up shot,
steam rising, soft morning light through a window
```

### Prompt Tips

<AccordionGroup>
  <Accordion title="Be specific about motion">
    Instead of "a bird", write "a hummingbird hovering in front of a red flower, wings beating rapidly". Motion descriptions help the model create more dynamic videos.
  </Accordion>

  <Accordion title="Describe camera work">
    Include camera directions: "tracking shot", "slow zoom in", "drone aerial view", "close-up", "pan left to right". This gives the video a professional, intentional feel.
  </Accordion>

  <Accordion title="Set the mood">
    Include lighting and atmosphere: "golden hour sunlight", "moody overcast", "dramatic shadows", "soft diffused light". These details significantly impact the final result.
  </Accordion>

  <Accordion title="Keep it focused">
    One clear subject and action per video produces better results than complex multi-character scenes. Start simple and iterate.
  </Accordion>
</AccordionGroup>

## Image to Video

Animate a still image into a video.

### How to Use

1. Open the **Image to Video** command
2. Provide one or more reference images (drag and drop, file picker, or URL)
3. Add a text prompt describing the desired motion and style
4. Generate and preview the result

### Best Practices for Reference Images

| Aspect          | Recommendation                                         |
| --------------- | ------------------------------------------------------ |
| **Resolution**  | High resolution produces better results                |
| **Subject**     | Clear, well-defined subjects animate better            |
| **Composition** | Center the main subject for predictable animation      |
| **Background**  | Simpler backgrounds allow more focus on subject motion |

### Local Reference Images

For xAI video workflows and other providers that require public media URLs,
EnConvo can upload a local reference image before sending the generation
request. You can choose an image from your Mac instead of manually uploading it
to a hosting service first.

<Tip>
  Use clear, high-resolution local images for image-to-video. If generation fails,
  try a smaller PNG or JPEG and confirm the file is fully downloaded from iCloud.
</Tip>

## Model Configuration

### Resolution (Sora Models)

| Resolution      | Aspect         | Best For                          |
| --------------- | -------------- | --------------------------------- |
| **720 x 1280**  | Portrait       | Social media stories, TikTok      |
| **1280 x 720**  | Landscape      | YouTube, presentations            |
| **1024 x 1792** | Tall Portrait  | Mobile wallpapers, vertical video |
| **1792 x 1024** | Wide Landscape | Cinematic, ultra-wide content     |

### Duration (Sora Models)

| Duration       | Use Case                                 |
| -------------- | ---------------------------------------- |
| **4 seconds**  | Quick clips, social media, previews      |
| **8 seconds**  | Standard clips, most use cases (default) |
| **12 seconds** | Longer scenes, storytelling              |

<Note>
  Duration and resolution options are currently available for Sora models. Other models (Kling, Hailuo, Veo, Wan) use their default output settings managed by the provider.
</Note>

## Cost and Points

### Enconvo Cloud Plan Pricing

| Model                          | Points per Video |
| ------------------------------ | ---------------- |
| **Kling Video v2.5 Turbo Pro** | 25,000           |
| **Hailuo 02 Standard**         | 22,500           |
| **Wan 25 Preview**             | 25,000           |
| **Veo3 Fast**                  | 37,500           |
| **Hailuo 02 Pro**              | 40,000           |
| **Sora-2**                     | 50,000           |
| **Veo3**                       | 100,000          |
| **Sora-2 Pro**                 | 150,000          |

<Tip>
  For the best value, start with **Kling Video v2.5 Turbo Pro** or **Hailuo 02 Standard** -- they offer excellent quality at the lowest point cost.
</Tip>

### Using Your Own API Keys

* **OpenAI (Sora):** Uses your OpenAI API balance. Pricing varies by resolution and duration.
* **Fal.ai:** Uses your fal.ai account balance. Supports Kling, Hailuo, Veo, and Wan models.
* **xAI:** Uses your xAI account and model access when xAI video features are available.

## Use Cases

<AccordionGroup>
  <Accordion title="Social Media Content">
    Create short, engaging video clips for Instagram Reels, TikTok, or YouTube Shorts. Use portrait resolution (720x1280) and 4-8 second duration for optimal social media fit.
  </Accordion>

  <Accordion title="Presentations and Demos">
    Generate visual demonstrations, concept animations, or background videos for presentations. Landscape resolution (1280x720) works best for slides.
  </Accordion>

  <Accordion title="Creative Projects">
    Explore concept art in motion, music video ideas, or short film scenes. Use Sora-2 Pro or Veo3 for the highest visual quality.
  </Accordion>

  <Accordion title="Product Visualization">
    Animate product photos into dynamic showcases. Use Image to Video with a product photo and describe the camera movement and environment.
  </Accordion>

  <Accordion title="Prototyping and Storyboarding">
    Quickly visualize scenes for video production planning. Generate multiple variations to explore different creative directions before committing to full production.
  </Accordion>
</AccordionGroup>

## Workflow Integration

Video generation integrates with other EnConvo features:

* **AI Chat:** Ask an AI agent to generate a video as part of a conversation
* **Workflows:** Chain video generation with other commands for automated content pipelines
* **File Management:** Generated videos are saved to your specified output directory with organized naming

## Limitations

<Warning>
  AI video generation is a rapidly evolving technology. Current limitations include:

  * Generation times range from 30 seconds to several minutes depending on the model
  * Complex multi-character scenes may have inconsistencies
  * Text rendering in videos is often inaccurate
  * Very specific motion sequences may not match your exact vision
  * All prompts should be in English for best results
</Warning>

## Troubleshooting

<AccordionGroup>
  <Accordion title="Video generation fails">
    1. Check your API key or Enconvo Cloud Plan balance
    2. Ensure your prompt is in English
    3. Try a simpler prompt -- overly complex descriptions can cause issues
    4. Switch to a different model to see if the issue is model-specific
    5. Check the console logs for error details
  </Accordion>

  <Accordion title="Video quality is poor">
    1. Use a more detailed, descriptive prompt
    2. Try a higher-quality model (Sora-2 Pro, Veo3, Hailuo 02 Pro)
    3. For Image to Video, ensure the reference image is high resolution
    4. Avoid requesting too many elements in a single scene
  </Accordion>

  <Accordion title="Generation takes too long">
    1. Switch to a faster model (Kling Turbo, Veo3 Fast, Hailuo Standard)
    2. Reduce the video duration if using Sora models
    3. Simplify the prompt -- fewer elements means faster generation
    4. Note that generation times depend on the provider's server load
  </Accordion>
</AccordionGroup>

## Related Features

<CardGroup cols={2}>
  <Card title="Image Generation" icon="image" href="/ai/image-generation">
    Generate still images from text
  </Card>

  <Card title="AI Chat" icon="comments" href="/ai/chat">
    Generate videos through AI conversation
  </Card>

  <Card title="SmartBar" icon="magnifying-glass" href="/features/smartbar">
    Quick access to video generation
  </Card>

  <Card title="Workflows" icon="diagram-project" href="/workflows/introduction">
    Automate video generation pipelines
  </Card>
</CardGroup>