Zhou Yucong leads MiniMax’s visual-model research and engineering and serves as an executive director.[1]
Start with MiniMax H3 introduction →
Education
Links
MiniMax biographyExecutive director · MiniMax
Visual-model research and engineering leader · MiniMax
Career
Zhou joined MiniMax in March 2022 and became an executive director in June 2025. He previously worked on algorithms at SenseTime (April 2018–July 2019) and Huawei (August 2019–March 2022). Since January 2023, he has also been Shanghai MiniMax’s legal representative and director.[1]
Contributions
MiniMax’s H3 release illustrates the company’s visual-model work. It combines text, images, video and audio as references for generating video with stereo sound. A prompt can specify a camera movement from one clip, a character from an image and vocals from an audio recording. The company describes training across generation and editing tasks together, rather than building separate systems for every task.[2]
Tools, writing & talks
Writing, research & talks
Start here
MiniMax H3 introductionminimax.ioMiniMax’s explanation and examples of multimodal video generation, using text, images, video and sound together as context.
Start with the reference-input and generated-video examples to see what multimodal generation means in practice. They make the visual-model program understandable before the architectural details.
Company release[2]