Search papers, labs, and topics across Lattice.
This survey critically examines the efficiency of GUI agents by analyzing their performance across various dimensions, including observation, memory, action, and runtime optimization. By synthesizing existing literature and identifying recurring mechanisms such as selective reading and hybrid runtimes, the authors highlight the importance of efficiency alongside task success for practical deployment. The study concludes with a discussion of open problems in the field, emphasizing the need for improved benchmarks and co-design strategies that address real-world constraints.
Efficiency in GUI agents is as crucial as task success, with recent advancements converging on innovative strategies like selective reading and hybrid execution models.
GUI agents increasingly operate across websites, mobile apps, and desktop environments, yet the field still reports progress primarily through task success. We argue that practical deployment depends equally on efficiency: how much context, computation, action budget, and runtime overhead an agent consumes while succeeding. This survey studies efficient GUI agents through an end-to-end systems lens that preserves the current technical axes of observation efficiency, context and memory efficiency, action efficiency, and planner-side/system efficiency. For each subsection, we expand the seed literature through targeted search plus backward and forward citation chaining, then synthesize the dominant mechanisms, reported efficiency signals, and new overheads they introduce. Across the literature, recent progress converges on a small set of recurring ideas: selective reading instead of full-context ingestion, global-to-local visual allocation, recoverable memory rather than raw history replay, verification-aware control, and hybrid runtimes that can switch between GUI and non-GUI execution. We conclude by identifying the main open problems, including honest accounting of verifier cost, cross-benchmark comparability, and co-design of observation, memory, and execution layers under real latency and privacy constraints.
