[MiniMax H3 Officially Released, Model Weights to Be Opened in the Coming Days] Deep Tide TechFlow reports that on July 31, MiniMax (00100.HK) officially announced the release of MiniMax H3. According to the introduction, this is a general-purpose, all-modal generative model that supports unified understanding of multimodal contexts composed of text, images, videos, and audio. It is capable of outputting native stereo audio and video, supporting up to 15 seconds at 2K resolution. MiniMax stated that they plan to open the model weights in the coming days, in compliance with relevant laws and regulations. (Jin10)