Jev-Mobile: Jev as an Executor for Mobile GUI Agents
Jev-Mobile splits VLM planning from fast GUI execution, cutting AndroidWorld API cost 73% at similar success.
Jev-Mobile separates low-frequency vision-language planning from high-frequency action execution for mobile GUI agents. The VLM specifies local goals, the accessibility tree defines the action space, and a fast typed model, Jev, selects actions so several steps follow one VLM decision. On the full AndroidWorld suite it reached 79% success, compared with 78% for SeeAct-V and 84% for a step-wise VLM. Among successful trajectories, mean runtime fell 32.7% and mean API cost fell 73.4% versus the step-wise baseline.
- A VLM sets local goals while Jev selects actions in the accessibility tree.
- AndroidWorld success was 79% for Jev-Mobile, 78% SeeAct-V, and 84% step-wise VLM.
- Successful runs cut mean time 32.7% and API cost 73.4% versus step-wise VLM.
Full article172 words · extracted from arxiv.org · click to collapse
Vision-language models (VLMs) have become a common foundation for autonomous mobile GUI agents, but most existing systems rely on the VLM for both planning and action grounding at nearly every interaction step, leading to substantial latency and model-serving cost. We introduce Jev-Mobile, which shifts this paradigm to low-frequency VLM planning and high-frequency lightweight execution: the VLM specifies local goals, the accessibility tree defines a structured executable action space, and Jev, a fast typed decision model, repeatedly selects actions within this space. This design allows multiple GUI actions to be executed under a single VLM decision, reducing expensive VLM inference while preserving adaptive interaction. On the full AndroidWorld task suite, Jev-Mobile achieves 79% task success, compared with 78% for SeeAct-V and 84% for a Step-wise VLM baseline. Among successful trajectories, it reduces mean end-to-end execution time by 32.7% and mean model API cost by 73.4% relative to Step-wise VLM. These results show that decoupling high-level VLM reasoning from low-level action execution can substantially improve mobile GUI agent efficiency while maintaining competitive task performance.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arxiv.org/abs/2609.30186