Photo + sound → video
Six small steps — pick a model, an image, an audio clip, the bit you want to use, write a prompt, choose camera, then render.
- 1Model
- 2Image
- 3Sound
- 4Sound part
- 5Prompt
- 6Camera
- 7Render
Choose the AI model
LTX-2.3 q8 is the recommended default — fast, fits a 24 GB GPU, and the one our prompts are tuned for. You can pick something else if you've downloaded it.
AI model