Every published Top AI Stories item tagged with MiniMax, newest first.
Days after an API-only launch, MiniMax put the weights for M3.1 Flash on Hugging Face: a multimodal mixture-of-experts model of about 428 billion parameters, 23 billion of them active per token, with a context window of up to one million tokens. The license is the MiniMax Community License, not an open-source one. Non-commercial use is free, while commercial users must display "Built with MiniMax M3.1-Flash" and notify the company, products earning more than 20 million dollars a year need written authorization, and military use is prohibited. The commercial thresholds match the ones MiniMax set for M3 in June.
Chinese lab MiniMax released M3.1 Flash, a fast model aimed at coding and agent work, through its API and its MiniMax Code tool. It costs 15 cents per million input tokens and 60 cents per million output tokens, and MiniMax says it cuts the cost of agent tasks by about 70 percent against its previous model. The company reports 73.7 percent on the DeepSWE coding benchmark and 61.7 on OSWorld V2, a test of operating a computer, but these figures are its own and the post does not name the models it compares against. MiniMax says the weights "will be released later" without giving a date or a license, so it is not yet an open-weight release.
A new Anthropic threat intelligence report attributes distillation campaigns with high confidence to Alibaba, Moonshot AI, DeepSeek, Zhipu, Xiaomi, SenseTime and MiniMax. A joint advisory from three US agencies asserted this pattern two days earlier; what is new here is the platform's own telemetry. Alibaba ran more than 151 million exchanges between May and July, peaking near three million a day across thousands of fraudulent accounts, to capture the chain-of-thought reasoning behind Opus. Moonshot and DeepSeek went further, silently routing their own customers' requests to Claude and keeping the transcripts, exposing names, live credentials and corporate data to a third party those users never chose.
The National Security Agency, the Federal Bureau of Investigation and the Cybersecurity and Infrastructure Security Agency issued a joint advisory saying DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z.ai have extracted billions of tokens from Claude, GPT, Gemini and Grok since late 2024, likely with Chinese government awareness. That escalates a July accusation from the White House technology chief, which named Moonshot alone. Distillation, training a smaller model on a larger one's outputs, is a standard technique — but the agencies say it forms the core rather than a supplement of these labs' development. They also reject the number behind the 2025 market shock, DeepSeek's widely cited training cost of 5.6 million dollars. The advisory says it leaves out the cost of the data taken this way. China's foreign ministry calls the claims groundless.
MiniMax launched H3 on July 31 and put the weights on Hugging Face on August 2, making it one of the few downloadable models that generates video with matching audio rather than silent clips. The 33 billion parameter system reads text, images, video and audio in a single context and returns 4 to 15 seconds at up to 2K resolution, 24 frames per second, with 32 kilohertz stereo sound. The MiniMax Community License is narrower than "open" suggests: it bars use of the weights outright in the United States, European Union, United Kingdom and South Korea, requires written authorization above $20 million in revenue, and mandates on-screen attribution. The hosted API is unrestricted.
Chinese lab MiniMax released M3, which it bills as the first open-weight model to combine frontier coding, agentic task execution, native multimodality, and a 1-million-token context window in one architecture. Built on a new sparse-attention design the team calls MSA, M3 scored 59 percent on SWE-Bench Pro — ahead of GPT-5.5 and Gemini 3.1 Pro and approaching Claude Opus 4.7. MiniMax says it will publish the open weights within ten days.