- Text-to-video generation
- Image-to-video generation (start frame, or start + end frames)
- Multi-shot video generation with auto or custom scene timing
- Audio generation
- Resolutions: 720p (std) and 1080p (pro)
- Aspect ratios: 1:1, 16:9, 9:16
- Duration: 3-15 seconds
Model
Parameters
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
prompt | string | ✅ Yes | - | Main text description for the video generation |
start_image_url | string | No | null | URL of the starting frame image |
end_image_url | string | No | null | URL of the ending frame image (ignored in multi-shot mode) |
duration | integer | No | 3 | Video duration in seconds (3-15). In custom multi-shot mode, this is calculated from shot durations |
mode | string | No | "std" | Mode: "std" (720p) or "pro" (1080p) |
aspect_ratio | string | No | "1:1" | Video aspect ratio: "1:1", "16:9", or "9:16" |
sound | boolean | No | false | Enable audio generation for the video |
enhance_prompt | boolean | No | false | Enable AI prompt enhancement |
multi_shot | array | No | null | Array of shots for multi-shot generation (see Multi-Shot Modes below) |
multi_shot_mode | string | No | "auto" | Multi-shot timing mode: "auto" or "custom" |
Multi-Shot Modes
1. Multi-Shot Disabled (Standard Generation)
Generate a single continuous video with one prompt. Example Request:2. Auto Multi-Shot Mode
Let the API automatically divide the duration among multiple shots. The system automatically calculates the duration for each shot to fit within the total duration. Example Request:- Set
multi_shot_modeto"auto" - Provide
multi_shotarray with prompts - API automatically divides the total
durationevenly among the shots - Each shot gets equal timing (e.g., 6 seconds total ÷ 2 shots = 3 seconds each)
3. Custom Multi-Shot Mode
Manually specify the exact prompt and duration for each shot. Full control over scene timing. Example Request:- Set
multi_shot_modeto"custom" - Each entry in
multi_shotarray must include bothpromptandduration - Total duration across all shots must not exceed 15 seconds
- The main
promptfield becomes optional (you can omit it entirely)
Multi-Shot Validation Rules
Custom Mode (multi_shot_mode: "custom")
- ✅ Each shot must have a
promptfield - ✅ Each shot must have a
durationfield (minimum 1 second) - ✅ Total duration of all shots must not exceed 15 seconds
- ✅ Each shot duration must be at least 1 second
Auto Mode (multi_shot_mode: "auto")
- ✅ Shots can have just prompts (durations auto-calculated)
- ✅ API will evenly distribute the total
durationamong shots
Resolution & Mode Mapping
| Resolution | Mode | Width | Height (1:1) | Width (16:9) | Height (16:9) | Width (9:16) | Height (9:16) |
|---|---|---|---|---|---|---|---|
720p | std | 720 | 720 | 1280 | 720 | 720 | 1280 |
1080p | pro | 1080 | 1080 | 1920 | 1080 | 1080 | 1920 |
Image Upload Support
When providingstart_image_url or end_image_url:
- Images are automatically uploaded and processed
- Start frame helps maintain character/scene consistency
- End frame guides the final pose/position (ignored in multi-shot mode)
- In multi-shot mode, only
start_image_urlis used
Examples
Example 1: Text-to-Video (Standard Mode)
Example 2: Image-to-Video with Start and End Frames
Example 3: Auto Multi-Shot Mode
Example 4: Custom Multi-Shot Mode
Response
Pricing
| Resolution | Price |
|---|---|
| 1080p | $0.04/s |
| 720p | $0.03/s |
| Quality | Extra Cost |
|---|---|
| Standard (720p) | +$0.075 |
| Pro (1080p) | +$0.12 |
