AI Glossary

Text-to-video

Also called T2V

What does “Text-to-video” mean?

Text-to-video: Generating a short video clip directly from a written description. The model must produce not just convincing frames but believable motion, camera movement and physics across time. Clips are typically seconds long, and some systems now generate synchronized audio as well.

Text-to-video visual guide: Generating a short video clip directly from a written description. The model must produce not just convincing frames but believable motion, camera movement and physics across time. Clips are typically seconds long, and some systems now generate synchronized audio as well.
Visual guide Open full size (new tab)

What is an example of “Text-to-video”?

OpenAI's Sora and Google's Veo are well-known examples; a prompt like "a drone shot over a misty forest at sunrise" yields a matching clip.

Keep learning

See also

Image-to-videoTermTemporal consistencyTermDiffusion modelTermText-to-imageTerm

Same topic

UpscalingTermText-to-speechTermVoice cloningTermStyle transferTermSpeech-to-textTerm