ltx2

To generate cinematic video clips from text prompts, you provide a descriptive text prompt to control framing, style, motion, and lighting. The Skill outputs a short, approximately five-second MP4 clip based on your specified resolution and frame rate.

Yes, you can add motion to a still image to create an animated video. The image-to-video feature adds subtle camera movement and ambient motion to portraits, slides, or screenshots, resulting in a short MP4 video clip.

Yes, text-to-video generation requires a Modal-hosted LTX-2.3 22B endpoint. You must have access to this environment to process prompt controls and generate the resulting MP4 video files.

You can create reproducible b-roll variations by adjusting prompt controls for frame count, resolution, fps, and quality. Using a specific seed value ensures the text-to-video or image-to-video output remains completely reproducible across generations.

AI generated video is best used for production workflows as b-roll for edits, animated backgrounds for title cards, cinematic tech visuals, and gentle motion for presentations. It removes the need for manual animation or expensive motion production for every cutaway.

The duration limitation for generating cinematic clips is that outputs are short, typically lasting around five seconds. You can adjust frame count and fps within the prompt controls, but the Skill is designed for short cinematic snippets rather than long videos.