千问办公上线首月用户数突破 3000万，企业用户占比过半
Ranking
No observed public metrics; popularity remains neutral/archived.
Merged summary
TL;DR — 阿里巴巴称，企业级 AI 智能体“千问办公”上线首月用户数突破 3000 万，其中企业用户占比过半。其增长依托企业系统集成、组织上下文基础设施和专用模型优化，旨在提升复杂办公任务的执行效率并降低成本。
- 千问办公可连接钉钉、CRM、ERP 等业务系统，将消息、文档、审批、邮件和知识库等异构数据转化为智能体可调用的组织上下文。
- 阿里开源了办公上下文基础设施 MyContext，用于处理海量异构工作数据；据称其在数周内获得超过 3000 个 GitHub Star。
- 专用 Qwen3.8-Flash 模型针对多步规划、工具选择和上下文压缩进行了调优，并结合推理及智能体运行框架优化。
- 据阿里披露，该方案在真实办公场景中使单任务生成速度提升一倍，平均 Token 消耗降低 75%。
- 产品已应用于汽车、金融、制造等行业；长安汽车称，某项工程计算的耗时由两天缩短至五分钟。
注：量子位更强调跨行业落地及长安汽车案例，雷峰网则更侧重产品快速迭代和整体工程优化。
Sources (2)
千问办公上线首月用户数突破 3000万，企业用户占比过半
TL;DR - Alibaba says its enterprise AI agent Qwen Office surpassed 30 million users in its first month, with enterprise users accounting for more than half. Its growth is paired with integrations across business systems and model-level optimizations aimed at faster, cheaper agentic workflows.
- Qwen Office connects with tools such as DingTalk, CRM, and ERP systems to turn messages, documents, approvals, email, and knowledge bases into actionable organizational context.
- Alibaba open-sourced MyContext, infrastructure for converting heterogeneous workplace data into agent-ready context; it reportedly gained more than 3,000 GitHub stars within weeks.
- A specialized Qwen3.8-Flash model was tuned for multi-step planning, tool selection, and context compression, reportedly doubling task-generation speed while reducing average token use by 75%.
- Enterprise deployments span automotive, finance, manufacturing, and other sectors; Changan Automobile reported reducing one engineering calculation from two days to five minutes.
千问办公上线首月用户数突破 3000万，企业用户占比过半
TL;DR - Alibaba says its Qianwen Office enterprise agent surpassed 30 million users in its first month, with enterprise users accounting for more than half. Its growth is paired with rapid product iteration, organizational-context infrastructure, and model-level efficiency improvements.
- The product integrates enterprise data from messages, documents, approvals, email, knowledge bases, CRM, and ERP systems into agent-callable context.
- Alibaba open-sourced MyContext, infrastructure for converting large volumes of heterogeneous workplace data into usable agent context; it reportedly gained over 3,000 GitHub stars within weeks.
- A specialized Qwen3.8-Flash model was tuned for multi-step planning, tool selection, and context compression, alongside inference and harness optimizations.
- Alibaba reports that this setup doubled per-task generation speed and reduced average token consumption by 75% in real office scenarios.