MiniMax宣布开源M3.1 Flash模型权重：428B参数、72小时处理Token超100万亿
北京——中国 AI 公司 MiniMax 10 月 23 日宣布，开源 M3.1 Flash 模型权重。据凤凰网科技报道，该模型总参数 428B（4280 亿）、激活参数 23B，采用混合专家（MoE）架构，上下文窗口达 100 万 Token。据 MiniMax 披露，M3.1 Flash 于 10 月 19 日上线，上线后 72 小时内真实用户处理 Token 量超过 100 万亿。
技术上，新模型将前三个全注意力层替换为稀疏注意力，并采用量化感知训练等三项优化提升推理效率。开源首日，模型已接入华为昇腾、平头哥、昆仑芯、摩尔线程、壁仞、海光、沐曦、天数、英伟达、AMD等芯片厂商，以及 Together AI、Fireworks、Nebius 等推理与云服务平台；vLLM、SGLang 及 HuggingFace、魔搭等社区同步推进适配。权重采用 MiniMax Community License：商用须署名，年营收超过2,000 万美元须事先取得 MiniMax 书面许可。
值得一提的是，此前以匿名身份"Space Bunny"在 OpenRouter 上处理超 200 万亿 Token 的神秘模型，其真身正是 M3.1 Flash 的早期版本。（GSM 据凤凰网科技、TechNews 科技新报综合）
来源：GSM·美国
MiniMax Open-Sources M3.1 Flash Model Weights: 428B Parameters, Over 100 Trillion Tokens Processed in 72 Hours
BEIJING — Chinese AI company MiniMax announced on October 23 that it is open-sourcing the weights of its M3.1 Flash model. According to Phoenix Tech, the model has 428 billion total parameters with 23 billion active, uses a mixture-of-experts (MoE) architecture, and offers a 1-milliontoken context window. MiniMax disclosed that M3.1 Flash launched on October 19 and processed more than 100 trillion tokens from real users within 72 hours.
Technically, the model replaces its first three full-attention layers with sparse attention and applies three optimizations, including quantization-aware training, to boost inference efficiency. On launch day, the model was already supported by chipmakers including Huawei Ascend, T-Head, Kunlunxin, Moore Threads, Biren, Hygon, MetaX, Iluvatar, NVIDIA and AMD, as well as inference and cloud platforms such as Together AI, Fireworks and Nebius; vLLM, SGLang, Hugging Face and ModelScope are advancing adaptation in parallel. The weights use the MiniMax Community License: commercial use requires attribution, and products or services with annual revenue above $20 million need prior written permission from MiniMax.
Notably, the mystery model that processed over 200 trillion tokens anonymously as "Space Bunny" on OpenRouter was revealed to be an early version of M3.1 Flash. (GSM, from Phoenix Tech and TechNews)
Source: GSM · America
来源：GSM·美国（原文链接）
图片来源：人工神经网络与芯片示意图（资料图）。照片：mikemacmarketing（原图发布于 flickr），CC BY 2.0，经 Wikimedia Commons。 / File photo. Photo: mikemacmarketing（原图发布于 flickr）, CC BY 2.0, via Wikimedia Commons.
