MMBench-GUI：面向多平台 GUI 智能体的统一分层评估框架
MMBench-GUI: A Unified Hierarchical Evaluation Framework for Multi-Platform GUI Agents
该基准将元素视觉定位（Element Grounding）确立为 GUI 智能体性能的关键瓶颈，为视觉定位研究提供了跨平台、细粒度的评测场景；其 EQA 指标同时衡量任务成功与动作冗余，可启发视觉智能体在效率与效果权衡上的评估方法设计。
原文依据 precise visual grounding is the critical determinant of performance
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出 MMBench-GUI，一个用于评估跨 Windows、macOS、Linux、iOS、Android 和 Web 的 GUI 自动化智能体的分层基准。该基准涵盖四个层级：内容理解、元素视觉定位、任务自动化和任务协作，覆盖 GUI 智能体的核心技能。为同时评估有效性和效率，我们进一步提出效率-质量感知（EQA）指标，在衡量任务成功的同时度量动作冗余。大量评估表明，精确的视觉定位是性能的关键决定因素，凸显了配备专用定位模块的模块化设计的优势。此外，所有智能体都存在显著的效率不足问题，即使最终成功也常常以过多步骤完成任务。在复杂或跨应用任务上性能也会下降，暴露出在记忆、规划和自适应推理方面的弱点。通过提供广泛覆盖、标准化协议和新颖指标，MMBench-GUI 为推进 GUI 智能体研究建立了首个全面基础。
英文原文摘要 1079 字符
We introduce MMBench-GUI, a hierarchical benchmark for evaluating GUI automation agents across Windows, macOS, Linux, iOS, Android, and Web. The benchmark spans four levels: Content Understanding, Element Grounding, Task Automation, and Task Collaboration, covering essential skills for GUI agents. To assess both effectiveness and efficiency, we further propose the Efficiency-Quality-Aware (EQA) metric, which measures task success alongside action redundancy. Extensive evaluations reveal that precise visual grounding is the critical determinant of performance, underscoring the advantages of modular designs with specialized grounding modules. Moreover, all agents suffer from substantial inefficiencies, frequently completing tasks with excessive steps despite eventual success. Performance also degrades on complex or cross-application tasks, exposing weaknesses in memory, planning, and adaptive reasoning. By providing broad coverage, standardized protocols, and novel metrics, MMBench-GUI establishes the first comprehensive foundation for advancing GUI agent research.