MiniMax AI 发布了其 MiniMax M3 模型，该模型具有开放权重，总参数约 4280 亿，激活参数 230 亿。该模型专为智能体时代设计，支持文本、图像和视频的原生多模态，上下文窗口高达 100 万 token。此次发布还包括更新的许可条款，年收入低于 2000 万美元的个人或公司可免费商用，只需通知。 AI
影响 这款具有大上下文窗口的开放权重多模态模型可能会加速智能体开发和多模态 AI 应用。
排序理由 前沿实验室模型发布，具有开放权重和系统卡。
AI 生成摘要 · Google Gemini · 来自 6 个来源。 我们如何撰写摘要 →
RT @gmi_cloud: MiniMax M3 is live on GMI M3 is the first open-weight model combining frontier coding & agent capabilities, 1M-token contex…
RT @novita_labs: 🤗 MiniMax M3 from @MiniMax_AI is now live on @huggingface — supported by Novita. Open weights. ~428B total parameters. ~2…
RT @RyanLeeMiniMax: On the M3 license — thanks for the feedback on M2.7 You told us prior-approval-for-any-commercial-use was too much. We…
RT @lmsysorg: 🎉 SGLang has Day-0 support for MiniMax-M3 from @MiniMax_AI, a native-multimodal MoE reasoning model of ~428B total params (~2…
MiniMax M3 is now available on Fireworks AI with Day-0 support — delivering the first open-weight model to combine frontier-level coding, native multimodal understanding (images and video), and a 1M-token context window in a single model. Powered by MiniMax Sparse Attention (MSA)…
RT @ModelScope2022: MiniMax M3 ist jetzt als Open Source verfügbar! Das Modell kombiniert natives multimodales Verständnis, ultra-lange Kontexte und Agent-Fähigkeiten in einem.🚀 mehr auf Arint.info # AI # DeepLearning # MachineLearning # MiniMaxM3 # ModelScope # OpenSource # arin…