With the advancement of multimodal large language models (MLLMs), building GUI agent systems has become an increasingly promising direction--especially for mobile platforms, given their rich app ecosystems and intuitive touch interactions. Yet mobile GUI agents face a critical dilemma: truly on-device models (4B or smaller) lack sufficient performance, while capable models (starting from 7B) are either too large for mobile deployment or prohibitively costly (e.g., cloud-only closed-source MLLMs). To resolve this, we propose OpenPhone, a mobile GUI agent system that leverages device-cloud collaboration to tap the cost-efficiency of on device models and the high capability of cloud models, while avoiding their drawbacks. Specifically, OpenPhone enhances Qwen2.5-VL-3B via two-stage SFT->GRPO training on synthetic GUI data for strong decision-making, integrates an efficient long-reasoning and memory management mechanism to utilize historical interactions under tight resources, and defaults to on-device execution--only escalating challenging subtasks to the cloud via real-time complexity assessment. Experiments on the online AndroidLab benchmark and diverse apps show OpenPhone matches or nears larger models, with a significant reduction in cloud costs.

翻译：随着多模态大语言模型（MLLMs）的发展，构建图形用户界面（GUI）智能体系统已成为一个日益前景广阔的方向——尤其是在移动平台上，鉴于其丰富的应用生态系统和直观的触控交互。然而，移动端GUI智能体面临一个关键困境：真正在设备上运行的模型（40亿参数或更小）性能不足，而能力足够的模型（从70亿参数起）要么对于移动端部署而言过大，要么成本过高（例如，仅限云端访问的闭源MLLMs）。为解决此问题，我们提出了OpenPhone，一个移动端GUI智能体系统，它利用设备-云端协同工作，以结合设备端模型的成本效益与云端模型的高性能，同时避免各自的缺点。具体而言，OpenPhone通过基于合成GUI数据的两阶段SFT->GRPO训练来增强Qwen2.5-VL-3B模型，以获得强大的决策能力；集成了一个高效的长推理与记忆管理机制，以便在资源受限条件下利用历史交互信息；并默认在设备端执行——仅通过实时复杂度评估，将具有挑战性的子任务升级到云端处理。在在线AndroidLab基准测试和多样化应用上的实验表明，OpenPhone达到或接近更大模型的性能，同时显著降低了云端成本。