MiniMax announced the open-sourcing of its MiniMax M3.1 Flash model weights, with 428B total parameters and 23B activated parameters. The model launched on October 19 and processed over 100 trillion tokens within 72 hours. MiniMax is partnering with chip vendors including Huawei Ascend, NVIDIA, and AMD, along with multiple cloud platforms, for day-one integration, with communities such as vLLM and Hugging Face supporting ecosystem adoption.
MiniMax released its new Flash model MiniMax M3.1 Flash, achieving generational improvements in agentic tasks, coding, computer use, and cybersecurity, with nearly 3x faster online inference and 70% lower usage costs. Input is priced at 1 yuan per million tokens, output at 4 yuan, and cache read at 0.1 yuan. It is now available on MiniMax API, MPlan, and MiniMax Code, with weights to be released later.
<body><div>Xiyu Technology open-sourced its native multimodal flagship model MiniMax M3, with 428B total parameters and 23B activated parameters, the first of its kind in</div></body>