MiniMax Launches H3 Model: Open-Source Multimodal AI Generates 2K Videos at One-Third Industry Cost

China's MiniMax releases H3 video model

BEIJING, July 31 (Reuters) - Chinese AI firm MiniMax (0100.HK), opens new tab released a new video-generation model on Friday that can process text, images, video and audio, stepping up competition in a fast-growing market led by rivals ByteDance and Kuaishou. The Shanghai-based company said its H3 model could generate videos of up to 15 seconds in 2K resolution with native stereo sound. It can also edit existing content and transfer movements from one video to another using instructions and reference material supplied in different formats. MiniMax said ⁠it planned to release H3's model weights within days, allowing users to download and customise the underlying system. The move would extend the open-weight approach increasingly adopted by Chinese AI developers into video generation, where many prominent models remain proprietary. Founded in 2022, MiniMax is one of China's "AI tigers", a group of well-funded startups competing with domestic tech giants and U.S. leaders in the crowded market for foundation models. It became the second of the group to list publicly when ⁠it floated in Hong Kong in January, following Z.AI (2513.HK), opens new tab. Reuters reported in early July, citing a person with direct knowledge, that MiniMax planned to launch H3 later in the month and that it was developing a 2.7-trillion-parameter language model. Competition among Chinese video-model developers has intensified since ⁠ByteDance released Seedance 2.0 earlier this year, drawing widespread attention for its ability to produce sophisticated clips from combinations of text, images, audio and video. Its launch followed Kuaishou's release ⁠of Kling 3.0. MiniMax said H3 was aimed at commercial applications including advertising, e-commerce, product design and games. It said generating 2K video would cost less ⁠than one-third of mainstream rival products. Amid a growing push in China's AI sector to reduce its dependence on U.S. semiconductors, MiniMax said H3 was designed to work with several Chinese-made chips. Reporting by Eduardo Baptista; Editing by Stephen Coates Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence Eduardo Baptista Thomson Reuters Eduardo Baptista is a Senior Correspondent for Reuters based in Beijing, covering China's technology, space, and automotive industries. He has led enterprise and investigative reporting on China's military-linked companies, artificial intelligence and semiconductor supply chains, as well as macroeconomic and industrial policy. Baptista has reported from China for nearly a decade and holds a BA in History from the University of Cambridge.

China's MiniMax launches H3 AI model; shares jump on multimodal push By Investing.com

Investing.com-- Shares of Chinese artificial intelligence startup MiniMax (HK:0100) jumped on Friday after the company unveiled H3, a new open-source multimodal AI model, intensifying competition in China's rapidly evolving AI sector. The company said H3 is a general-purpose multimodal generation model capable of processing unified context across text, images, video and audio, while generating videos with native stereo sound of up to 15 seconds in 2K resolution. Get real-time updates on market-moving news with InvestingPro -- now at 60% off Hong Kong-listed shares of the company surged 17.4% to HK$230.0 by 04:24 GMT. MiniMax said the model is aimed at commercial applications including advertising, e-commerce, gaming, product design and content creation. MiniMax said H3 delivers industry-leading price-performance, with 2K video generation costing less than one-third of comparable mainstream models on a per-second basis. At 768p resolution, it said pricing is less than half that of competing models generating 720p video. The company also plans to release the model's weights in the coming days, subject to applicable laws and regulations, to support developers and improve compatibility across AI hardware platforms. The launch comes as Chinese AI firms race to release increasingly capable open-source models to challenge U.S. rivals. MiniMax said H3 introduces technologies including Contextual Omni Representation, the H3-Omni Transformer and In-Context Regeneration, enabling stronger instruction following, multimodal understanding and high-resolution content generation.

Why is MiniMax stock surging today? By Investing.com

Investing.com -- MiniMax stock surged 15.1% on Friday after the company officially released MiniMax H3, its next-generation multimodal generative AI model. The release marks one of the most significant product launches in the company's short history as a publicly traded firm. H3 is designed to generate video content up to 15 seconds in length at native 2K resolution with built-in stereo audio, accepting text, image, video, and audio as inputs, with Minimax also outlining plans to open-source the model within days. The company also said the model took up approximately one-third of prevailing industry costs, a pricing point that analysts and market participants viewed as a meaningful competitive differentiator. Adding to the momentum, the launch coincided with rising investor expectations that MiniMax is on the verge of joining the Stock Connect program, which would allow mainland Chinese investors to purchase the stock directly through the southbound trading link. Estimates had previously suggested that Stock Connect eligibility could bring tens of billions of Hong Kong dollars in fresh inflows into the stock, creating a structural demand catalyst that amplified the market's reaction to the H3 news. The broader market environment provided a supportive backdrop. Asian tech and AI stocks rallied hard on Friday, tracking an overnight jump in Wall Street following upbeat earnings from Microsoft and Amazon. This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.