LTX-2 Text-to-Video API
Generate professional-quality videos from text descriptions using the LTX-2 API Text-to-Video endpoint. Learn prompt engineering techniques and best practices.
Overview
The Text-to-Video (T2V) endpoint transforms natural language descriptions into high-quality video content. Simply describe what you want to see, and the LTX-2 model generates a corresponding video clip.
Endpoint: POST /v1/text-to-video
Prompt Engineering Tips
The quality of your generated video depends largely on how well you craft your prompt. Here are key techniques:
1. Be Specific and Descriptive
Include details about subjects, actions, environment, lighting, and style.
2. Describe Camera Movement
Specify how the camera should move for dynamic videos.
- • "Camera slowly pans left to right"
- • "Drone shot rising above the landscape"
- • "Close-up tracking shot following the subject"
- • "Static wide shot"
3. Include Style Keywords
Add style modifiers to achieve specific looks.
4. Specify Temporal Elements
Describe how scenes should evolve over time.
- • "Time-lapse of clouds moving across the sky"
- • "Slow-motion water droplets falling"
- • "Day transitioning to night"
Example Prompts
"Aerial drone shot of a winding coastal road at sunset, waves crashing against cliffs, golden hour lighting, cinematic 4K quality"
"Close-up of a steaming cup of coffee on a wooden table, morning light streaming through window, cozy cafe atmosphere, shallow depth of field"
"Futuristic cityscape at night with flying cars and neon signs, camera slowly tilts up to reveal massive skyscrapers, cyberpunk style, rain reflections on streets"
API Request Example
{
"prompt": "A majestic eagle soaring over snow-capped mountains, golden hour sunlight, cinematic quality, camera tracking the eagle's flight path",
"negative_prompt": "blurry, low quality, distorted",
"duration": 10,
"resolution": "1080p",
"fps": 25,
"mode": "pro",
"seed": 42,
"webhook_url": "https://your-server.com/webhook"
}