ByteDance OmniHuman v2
bytedance/omnihuman/v2Animate a human figure from one image and a WAV or MP3 audio track. Add an optional prompt to guide the scene, movement and camera. Audio must be 4–300 seconds and at most 15 MB. Images must be JPEG, PNG or WebP, at most 30 MB, 300–6000 pixels on each side, with aspect ratio 0.4–2.5. Video length follows the audio. Prompts support Chinese, English, Japanese, Korean, Mexican Spanish and Indonesian.
BytePlus Video Enhancement Pro
bytedance-video-enhancement-proApply stronger BytePlus video restoration and enhancement for footage with real people. Pro improves skin texture and fine detail while supporting upscaling to 2K or 4K, frame interpolation up to 120 FPS, denoising, deblocking, deblurring, sharpening, and color and contrast enhancement.
BytePlus Video Enhancement Standard
bytedance-video-enhancement-standardPost-process videos with BytePlus restoration and enhancement. Upscale to 2K or 4K, interpolate up to 120 FPS, repair noise, blocking, blur, scratches, and jitter, and improve color and contrast across AI video, UGC, short drama, and archival footage.
MiniMax H3 Highspeed
minimax/h3-highspeedMiniMax H3 Highspeed creates 4–15 second videos at 768p with native audio. Generate from text, animate a starting image with an optional end frame, or combine up to nine image, three video, and three audio references. Reference clips must be 2–15 seconds each, totaling at most 15 seconds per media type. Audio references require an image or video. Frame guidance and reference inputs cannot be combined.
