AI Glossary / Text-to-video

🎨 Image, video and audio

Text-to-video is generating moving footage from text descriptions, and increasingly also from a single image (image-to-video), which the model animates. Contemporary systems create clips from a few to several seconds long with believable motion physics, and the best generate synchronized sound too. The biggest challenges are character consistency between shots, longer sequences, and precise directorial control. In practice, creators treat these tools as factories for individual shots to edit together, not machines for finished films.