Video Generation
Create AI-generated videos from text prompts.
Stable
Overview
elizaOS Cloud provides video generation through Fal.ai, supporting multiple models:
- Google Veo 3: Highest quality, latest model
- Kling Video: Multiple tiers from standard to master
- MiniMax Hailuo: Fast and affordable
Quick Start
Dashboard
Navigate to Dashboard → Videos for the visual interface.
API
cURL
curl -X POST "https://cloud.milady.ai/api/v1/generate-video" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "A butterfly emerging from a cocoon in slow motion",
"model": "fal-ai/veo3"
}'Response
{
"success": true,
"video": {
"url": "https://your-storage.vercel-storage.com/videos/abc123.mp4"
},
"model": "fal-ai/veo3"
}Available Models
| Model | Provider | Quality | Speed |
|---|---|---|---|
| `fal-ai/veo3` | Highest | Slow | |
| `fal-ai/veo3/fast` | High | Fast | |
| `fal-ai/kling-video/v2.1/master/text-to-video` | Kling | Highest | Slow |
| `fal-ai/kling-video/v2.1/pro/text-to-video` | Kling | High | Medium |
| `fal-ai/kling-video/v2.1/standard/text-to-video` | Kling | Good | Fast |
| `fal-ai/minimax/hailuo-02/pro/text-to-video` | MiniMax | High | Medium |
| `fal-ai/minimax/hailuo-02/standard/text-to-video` | MiniMax | Good | Fast |
Request Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
| `prompt` | string | ✓ | Text description of the video |
| `model` | string | Model to use (default: `fal-ai/veo3`) |
Prompt Tips
Camera Movements
Include camera direction for dynamic videos:
- `camera pan left/right` - Horizontal movement
- `zoom in/out` - Focal adjustment
- `dolly forward/backward` - Physical camera movement
- `tracking shot` - Following subject
- `aerial view` - Bird’s eye perspective
Scene Description
Be specific about motion and timing:
{
"prompt": "A golden retriever running through autumn leaves in a park, slow motion, cinematic lighting, tracking shot following the dog"
}Style Modifiers
- `slow motion` - Dramatic effect
- `timelapse` - Accelerated time
- `cinematic` - Movie-like quality
- `handheld` - Organic movement
- `steady cam` - Smooth, professional
Best Practices
- Be specific about motion: Describe what moves and how
- Include timing cues: Slow motion, timelapse, etc.
- Specify camera work: Pan, zoom, tracking shots
- Set the mood: Cinematic, documentary, etc.
Video generation typically takes 30-180 seconds depending on model and complexity.
Error Handling
| Code | Error | Solution |
|---|---|---|
| 400 | Invalid prompt | Ensure prompt is non-empty |
| 400 | Invalid model | Use a supported model |
| 402 | Insufficient credits | Add credits to your account |
| 429 | Rate limited | Wait and retry |
| 500 | Generation failed | Retry or try a different model |