Generate a video in LTX Desktop
LTX Desktop supports text-to-video, image-to-video, audio-to-video, Retake, and gap fill — all from the Gen Space panel. Here's how each mode works.
Text-to-video
- Open LTX Desktop and go to Gen Space.
- Type your prompt in the text field.
- Choose your resolution, duration, and model.
- Click Generate.
The generated clip appears in your asset bin when complete. From there you can add it to the timeline or generate additional takes.
Image-to-video
- In Gen Space, click the image input area and upload a reference image.
- Add a prompt describing how you want the image to animate.
- Set your duration and click Generate.
Audio-to-video
- In Gen Space, make sure the mode is set to Video.
- Click the music note icon in the prompt area and attach an audio file.
- Optionally add a text prompt to guide the visual style.
- Click Generate.
Retake
Retake lets you regenerate a specific section of a clip while keeping the surrounding frames consistent.
- Hover over a clip in your asset bin and click Retake, or select Retake from the Mode dropdown in Gen Space.
- Trim the section you want to redo in the Retake panel.
- Adjust your prompt if needed.
- Click Retake.
Note that Retake in API mode consumes API credits.
Gap fill
Gap fill generates a clip that bridges two existing clips in your timeline, matching the motion and content at both ends.
- Place two clips in the timeline with a gap between them.
- Select the gap.
- Click Gap fill and add a prompt if needed.
- Click Generate.
Multiple takes
Each generation creates a new take. You can switch between takes for any clip non-destructively in the timeline — earlier takes are preserved until you delete them.
If your machine uses API mode, video generation and Retake consume LTX API credits. Local generation on supported hardware has no per-generation cost.