AI 导读
MiniMax 发布 MiniMax M3，官方称其为首个同时具备编码与智能体、1M 上下文和原生多模态三项前沿能力的开放权重模型。该模型在 SWE-Bench Pro 得分 59.0%、Terminal Bench 2.1 得分 66.0%、MCP Atlas 得分 74.2%，并通过 MiniMax Sparse Attention 将上下文扩展到 1M。API 已在 platform.minimax.io 上线，权重与技术报告预计约 10 天后发布，转发者表示期待上手测试其在 OpenHands 中的表现。
正文
Looking good 👀
Historically MiniMax models have worked well in OpenHands, looking forward to giving this one a whirl!
Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities - Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas - MiniMax Sparse Attention scales context to 1M - Natively Multimodal from Step Zero API: platform.minimax.io Token Plan: platform.minimax.io/subscrib… 🚀New! MiniMax Code: code.minimax.io Weights & Tech Report in ~10 Days在 X 查看被引用的帖子
来源：@gneubig · x.com