video

AI prompt gallery and market for image/video generations.

Full-stack text/image/video moderation plus deepfake & AI-content detection.

Professional CV labeling with model-in-the-loop and medical/video tooling.

Open multimodal family (1B–78B); 78B surpasses 70% on MMMU; broad image/video understanding.

Fast voice cloning for video narration and multilingual dubbing; pairs with avatars and lip-sync.

Clone your voice in minutes and generate narration inside Descript’s editor for podcasts and video.

Create narration and videos from scripts with 800+ voices across 100+ languages.

Batch background removal and background changer for photos and videos.

Remove image backgrounds online; also supports smart removal in videos.