Lightricks LTX 2.3 Text To Video

Generate cinematic video clips from text prompts with controls for framing, duration, frame rate, resolution, and audio.

Model Overview

LTX 2.3 Text-to-Video creates short video clips directly from prompt text. It exposes core generation controls so you can tune framing and playback characteristics for different output targets.

Best At

• Prompt-driven short-form cinematic video generation.
• Fast creation of story beats and visual concepts.
• Controlled outputs with explicit aspect, resolution, and timing settings.

Limitations / Not Good At

• Output quality depends on prompt specificity and structure.
• Very complex scene logic may require iterative prompt refinement.
• Longer or denser prompts can introduce variability across runs.

Ideal Use Cases

• Storyboard and concept-visual prototyping.
• Social video draft generation from creative briefs.
• Rapid campaign visual exploration from text direction.

Input & Output Format

• Input: required prompt; optional aspect_ratio, resolution, duration, fps, and generate_audio.
• Output: generated video URI returned on response.

Performance Notes

• Clear subject/action/camera wording improves first-pass quality.
• Use conservative parameter combinations for stable iterative loops.
• Enabling audio can improve completeness for end-user previews.