Text to video: generate a clip from a prompt

tendre.AI turns a text prompt into a short video clip with real motion and synchronized sound, powered by LTX-2.3. Everything renders on your own GPU in Pro mode, or in the cloud in Starter mode if you prefer.

1. Open the Video tab

Switch to the Video tab and choose Text to video mode, as opposed to Image to video or Transition.

2. Describe the scene

Write a prompt describing the subject, the action and the setting, much like an image prompt but with movement in mind, for example "a woman walks through a neon-lit street, camera slowly follows." Add a short note on the sound too, such as ambient city noise, if you want one.

3. Set length and sound

Choose a clip length between 5 and 20 seconds and confirm the sound toggle is on. LTX-2.3 generates audio synchronized to the motion in the clip, not a track added afterward.

4. Generate

Press Generate. Local renders use your GPU, so timing depends on its power and the chosen length, with a progress bar tracking the render. Longer clips and higher resolutions take proportionally more time.

5. Prefer the cloud

If your GPU is busy or not powerful enough, switch to Starter mode from Settings to render the same clip in the cloud instead, billed in credits.

See Image to video to animate a still instead of starting from a prompt, and Camera movements to add a cinematic move to the clip.