Text-to-video
Also: t2v
Generating a short video clip from a text prompt with a diffusion model. It is far more memory- and compute-intensive than image generation, and open models are only recently consumer-runnable.
Text-to-video extends diffusion from a single image to a sequence of frames that move coherently. That coherence is expensive: video models are much larger and far more demanding on memory and compute than image models, which is why open text-to-video has, until recently, needed data-centre hardware.
That is now starting to change. Models like Wan 2.2 fit a single 24GB card, while higher-quality options such as HunyuanVideo still want 45GB or more. Clips are typically only a few seconds long. Many models also do image-to-video, animating a still image.