Skip to content
Monster社区 · 中文ComfyUI实测与工作流·· 21 days agoAI score50

MiniMax H3 2K Refiner: Separation of the Upscaling Module and Audio-Visual Latent Space

AI briefing

AIwood Aiwu Research Lab demonstrated the MiniMax H3 2K Refiner workflow, connecting a 2K upscaling module after the existing generation pipeline, and separating the audio-visual latent space to avoid redrawing affecting the relationship between audio and lip sync.

Open original bbs.monster