⏱️OS-Marathon

Motivation

😆 🧳

🤦 💧

Traveling receipts

Expense system

🤔 Can computer-use agents help to do these long tasks?

Vast-Horizon, Repetitive Tasks

📋 TASK:

Do an expense report for me

⚙️ Decompose the task

🤖 ✨ ⭐ 💫

Receipts

✓

✓

✓

✓

...

📁 Sub-Workflow 1
(Data Unit 1)

🔍 Identify the required data

📖 Understand and find the required information from data

✏️ Fill the form

Processing Receipt

READING

OS-Marathon targets vast-horizon, repetitive tasks.

✨ News & Updates

• 21/08/2026: Our paper is accepted to EMNLP 2026 Main Conference! We are cleaning the code, which will be released soon.
• 1/29/2026: Paper released on arXiv.
• 1/28/2026: Website released.

💡 Key Takeaways

We identify three key challenges for computer-use agents on vast-horizon, repetitive tasks:

📑 Abstract

Vast-horizon, repetitive workflows are common in daily routines, e.g., processing expense reports from a stack of receipts, organising a collection of PDF annotations into structured notes, and are tedious for humans, with execution length scaling with the volume of data to process. They are, however, well-suited to autonomous agents, since their structured, recurring sub-workflows follow a logic that can be systematically learned. Yet no existing benchmark evaluates agents specifically on vast-horizon, repetitive capabilities, leaving the problem largely underexplored. To bridge this gap, we establish OS-Marathon, comprising 100 vast-horizon repetitive tasks across 5 scenarios and 10 domains, on which our evaluation reveals that leading state-of-the-art (SOTA) CUAs struggle substantially. We further find that introducing a task orchestrator to decompose the workflow into per-instance subtasks fails to mitigate the challenge: errors accumulate and propagate across solver agents, indicating that naive decomposition is insufficient for these tasks. We then explore a cost-friendly personalisation strategy, i.e. GraphDemo, that adapts general agents to such tasks from a single human demonstration of the recurring sub-workflow logic. Extensive experiments show both the real-world challenge of vast-horizon, repetitive tasks and the improvement provided by human demonstration in this setting.

🚀 Task Workflow Visualization

We visualize the execution process of two tasks from the expense report and transcript domains to illustrate the complete workflow. Use the controls to switch between examples.

📊 Statistics of the Benchmark

🖥️ Execution Environment Visualization

We visualize the execution environment for the tasks. Note that this illustration is for conceptual purposes only, and the data shown may differ from the data used in practice.

University Expense System Environment

Corporate Expense System Environment

GPA Calculator System Environment 1

GPA Calculator System Environment 2

Corporate Expense Spreadsheet Environment 1

Corporate Expense Spreadsheet Environment 2

University Expense Spreadsheet Environment

Previous Next

✏️ Qualitative Results

We provide the qualitative results of some baseline CUAs on our benchmark. We attach the agents' action prediction on the top right corner of the video.

📐 Quantitative Results

We compare the performance of the agents on our benchmark. We evaluate the form field-level and subtask-level success rate of the agents across the 5 scenarios in OS-Marathon.

🙏 Acknowledgements

This work was conducted during an internship at Microsoft; we thank Microsoft Research (MSR) and Windows Cloud Experience (WCX) for their support. We also thank the authors of the OSWorld benchmark for their open-source infrastructure, which served as a foundation for this project.

📚 Citation

@article{wu2026OS-Marathon,
        title={OS-Marathon: Benchmarking Computer-Use Agents on Vast-Horizon Repetitive Tasks},
        author={Wu, Jing and Ai, Wenjie and Barretto, Daphne and Chen, Yiye and Qingyu, Chen and He, Yuhang and Pranit, Chawla and Gydé, Nicholas and Jian, Yanan and Vineet, Vibhav},
        journal={EMNLP Main Conference},
        year={2026}
}