Open Access

ARTICLE

ComAlign: A Benchmark Aligning Natural Language with Operating System Commands

Shasha Li, Bin Ji^*, Xiaodong Liu, Jun Ma, Jie Yu^*

College of Computer Science and Technology, National University of Defense Technology, Changsha, China

* Corresponding Authors: Bin Ji. Email: ; Jie Yu. Email:

Computers, Materials & Continua 2026, 87(2), 54 https://doi.org/10.32604/cmc.2026.076083

Received 13 November 2025; Accepted 29 December 2025; Issue published 12 March 2026

Abstract

Aligning natural language with operating system (OS) commands allows users to perform complex computer tasks through simple natural language descriptions. However, due to the complex nature of natural language, it still remains challenging to achieve precise alignment. In this paper, we present ComAlign, a Chinese benchmark dataset that pairs Chinese natural language descriptions with corresponding OS commands. ComAlign covers a broad range of 82 distinct OS command types with a total of 1811 natural language descriptions. We elaborate on the construction of ComAlign and construct three baselines to evaluate the alignment accuracy on ComAlign. Experimental results show that even advanced large language models struggle with certain ambiguously phrased OS commands. Specifically, the best performing baseline achieves 46.9% alignment accuracy. We demonstrate that ComAlign is collected from real-world application scenarios, making it particularly suitable for developing and benchmarking intelligent OS and agent systems that support user-machine interactions through natural language.

---

Keywords

---