CutVerse: A Compositional GUI Agents Benchmark for Media Post-Production Editing

TL;DR

CutVerse introduces a comprehensive benchmark for evaluating autonomous GUI agents in professional media post-production tasks, highlighting current limitations and guiding future research in complex, multimodal workflows.

Contribution

The paper presents a new benchmark with expert demonstrations and a parser for structured evaluation of GUI agents in media editing, addressing a gap in autonomous agent capabilities.

Findings

Existing agents achieve only 36.0% success on complex tasks

Current models show promise in spatial grounding and multimodal alignment

Long-horizon reliability remains a significant challenge

Abstract

While GUI agents have made significant progress in web navigation and basic operating system tasks, their capabilities in professional creative workflows remain largely underexplored. To bridge this gap, we introduce Cutverse, a benchmark designed to systematically evaluate autonomous GUI agents in realistic media post-production environments. We curate expert demonstrations across 7 professional applications (e.g., Premiere Pro, Photoshop), covering 186 complex, long-horizon tasks grounded in authentic editing workflows, involving dense multimodal interfaces and tightly coupled interaction sequences. To support scalable evaluation, we develop a lightweight parser that transforms raw screen recordings and low-level interaction logs into structured, compositional GUI action trajectories with precise grounding. Extensive evaluations reveal that existing agents achieve only 36.0\% task success on realistic media editing tasks, underscoring the challenges posed by complex, long-horizon media post-production workflows in our benchmark.While current models demonstrate promising spatial grounding, multimodal alignment, and coordinated action execution, they remain limited in long-horizon reliability and domain-specific planning.