MiniMax plans China's biggest AI model with 2.7 trillion parameters, will open-source it

MiniMax plans China's biggest AI model, and will open-source it

MiniMax is building a 2.7-trillion-parameter model, six times the size of its current flagship and the largest in China, The Information reports. It plans to open-source the system, targeted for the third quarter, in another blow to the paid frontier. MiniMax plans its biggest model yet, a 2.7-trillion-parameter system it intends to open-source, The Information reports. It would be the largest Chinese model on the market, and another squeeze on US labs' margins. China's open-model push has a new heavyweight on the way. MiniMax plans a large language model with 2.7 trillion parameters, The Information reports, citing two people familiar with the effort. It would dwarf the firm's current flagship and rank as China's largest model yet. Inside MiniMax, the model goes by M3 Pro. The final name may change. The company aims to open-source it, and a release could land as early as the third quarter. The scale of the jump is the headline. MiniMax's current flagship, M3, holds 428 billion parameters. The new model runs roughly six times bigger. Bigger models tend to handle complex reasoning and multi-step tasks better, which is exactly where the frontier race now sits. A crowded Chinese field MiniMax does not run alone. It faces fierce competition from DeepSeek, Zhipu and Moonshot AI. Each races to ship capable models cheaply, and many carry open weights. The stakes are commercial as well as technical. MiniMax trades publicly in Hong Kong, where AI stocks have swung hard this year. A strong open model would help it stand out in a brutal home market. Cheap, open, and everywhere The bigger story sits outside China. Chinese open models have spread fast in 2026. Developers reach for them to run high-volume, less-critical work at a fraction of the price of US frontier systems. Enterprises increasingly adopt open models too. That worries the American labs. They spend billions to train frontier models, then charge a premium. Cheap, capable open weights undercut that logic. The friction already spills into fights over distillation and access. Beijing, for its part, weighs curbs on who abroad can use its best models. Why it matters A 2.7-trillion-parameter open model would make a statement. It would hand developers worldwide a free, powerful option, and pile more pressure on the paid frontier. For MiniMax, the launch doubles as a test. Win, and it pulls clear of its Chinese rivals. Fall short, and the open-weight crown stays up for grabs. The proof comes when the weights drop, perhaps this quarter.

China's MiniMax plans to launch giant 2.7 trillion parameter model

Chinese AI firm MiniMax is developing a 2.7 trillion-parameter model. This new model could become the world's largest open-weight AI system. Cheaper Chinese AI models are gaining traction as alternatives to US systems. MiniMax will also launch a multimodal video generation model soon. The company is planning a second listing on Shanghai's STAR Market. Chinese AI startup MiniMax Group is working on a large language model with 2.7 trillion parameters, a person with direct knowledge told Reuters, the largest open-weight AI model released by a Chinese firm and possibly the largest in the world. The global push toward trillion-parameter AI models is being driven by growing demand for ⁠autonomous systems ⁠capable of complex reasoning. Analysts say the threshold is critical to building systems that can execute multi-step operations without human intervention. The new model could be released as early as the third quarter, the person said, declining to be identified as the information is not public. The company declined to comment. Lower-cost alternatives to US systems: Plans for the colossal model - details of which were first reported by The Information - ⁠come as cheaper, open-source-based models from Chinese providers such as MiniMax, Z. ai and DeepSeek are gaining traction in the U.S. and other hubs as lower-cost alternatives ⁠to proprietary U.S. systems. Open-weight models allow users to download, run and customise the underlying systems, unlike proprietary, closed-source models. MiniMax will later this month launch H3, its frontier-level multimodal video generation model, the source added. Founded in 2021, MiniMax is an upstart in China's AI industry, raising HK$4.8 billion ($614 million) in its Hong Kong initial public offering in January. It is planning a second listing in Shanghai's tech-heavy exchange, the STAR Market. Meituan's LongCat-2.0 and DeepSeek's V4-Pro lead China's AI industry with 1.6 trillion total parameters, while several other domestic rivals have passed the trillion-parameter threshold. While standard generative chatbots excel at short-form processing, they hit ⁠mathematical limitations when tasked with independent, long-horizon decision making. Mixture of experts engineering: This massive architectural expansion relies heavily on Mixture of Experts engineering to balance intelligence with operational costs. By organizing a model into specialized subnetworks, developers can build vast, trillion-parameter databases that only activate a small fraction of their capacity per query. The approach gives users access to deeply specialized domain knowledge, including complex legal codes and rare software bugs, at the high speed and lower cost of mid-sized systems.