Quick start
Here’s a simple example using Lucy-Pro, our highest quality video model.Generate from text
Animate images
Control object motion with trajectories
For precise control over object motion, use Lucy Motion with trajectory paths instead of prompts:data(required) — Input image file to animatetrajectory(required) — Array of keyframe coordinates defining motion path:[{frame: number, x: number, y: number}]seed(optional) — Random seed for reproducibilityresolution(optional) — Output resolution:720p(default)
Parameters
Text to Video
prompt(required) — Text description of the video to generate.resolution(optional) — Output resolution:480por720p(default:720p).
Image to Video
data(required) — Input image file to animate.prompt(required) — Text description of how to animate the image.resolution(optional) — Output resolution:480por720p(default:720p).
For complete API documentation including response formats and error codes, see the API Reference for Text to Video and Image to Video.
Available Models and Options
Decart provides different configurations to balance speed, quality, and cost depending on your use case. We currently offer three model families:- Lucy-Pro — our flagship model for generating diverse, high-quality videos with prompts.
- Lucy-Motion — trajectory-based motion control for precise, predictable animations.
- Lucy-Dev — a faster model ideal for quick prototyping and simple animations.
Video specifications:
- Duration: Fixed 5 seconds
- Resolution:
480p: 480×832 (portrait) or 832×480 (landscape)720p: 720×1280 (portrait) or 1280×720 (landscape)
- Format: MP4 with H.264 encoding
Prompt Engineering
For Text-to-Video
Video generation prompts should describe both what happens (subjects, actions) and how it looks (camera movement, style, atmosphere). Because video generation involves temporal elements, your prompts should consider movement, pacing, and progression. Effective prompts are specific about motion (around 80–100 words), include camera instructions (pan, zoom, static), and describe temporal progression (beginning, middle, end states). Lucy handles cinematic language well—camera angles, lighting changes, and atmospheric effects. Example prompts:- “A majestic eagle soaring through mountain valleys, camera tracking its flight. Golden hour lighting with dramatic clouds. The eagle dives toward a lake, then rises with a fish. Cinematic aerial perspective.”
- “Time-lapse of a busy city intersection from sunrise to sunset. Traffic flows like rivers of light, shadows grow and shrink, sky transitions from pink to blue to orange. Static wide angle shot.”
Endpoints and Full API Reference
- Lucy-Pro Text-to-Video:
/v1/generate/lucy-pro-t2v - Lucy-Pro Image-to-Video:
/v1/generate/lucy-pro-i2v - Lucy-Dev Image-to-Video:
/v1/generate/lucy-dev-i2v - Lucy-Motion Image-to-Video:
/v1/generate/lucy-motion