paper-with-me

Papers

TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents

2026-06-26 · Shoufa Chen, Luyuan Wang, Xuan Yang, Zhiheng Liu, Yuren Cong, Yuanfeng Ji, Feiyan Zhou, Xiaohui Zhang, Fanny Yang, Belinda Zeng arxiv

As large language models and harness frameworks continue to advance, agents operating in terminals are increasingly capable of performing a broader range of general computer-use tasks beyond coding. However, existing benchmarks do not adequately evaluate general-purpose terminal computer-use agents (TUAs): general computer-use benchmarks primarily target graphical user interfaces (GUIs), whereas terminal-based benchmarks largely emphasize technical and programming-centric workflows historically native to the shell. We introduce TUA-Bench, a general-purpose benchmark for terminal-use agents. TUA-Bench includes 120 real-world tasks across five task families, covering routine digital activities-including document editing, email management, and live-web information seeking-as well as scientific and engineering workflows co-designed with PhD-level domain experts that require specialized software. This breadth distinguishes TUA-Bench from prior shell-focused or domain-specific benchmarks. Each task is manually designed, runs in a real terminal with a deterministic setup script, and is evaluated by an execution-based scoring protocol. We find that the strongest frontier agent, Claude Code with Claude Opus 4.8 max reasoning effort, achieves 65.8% overall performance, with substantial gaps across both tracks. By providing a broad and realistic evaluation of terminal-use capabilities, TUA-Bench aims to accelerate the transition from narrow, task-specific assistants to general-purpose agents capable of operating reliably across diverse digital environments.

📄 PDF Abstract BibTeX arXiv:2606.28480

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks

2026-05-21 · Zhaoyang Chu, Jiarui Hu, Xingyu Jiang, Pengyu Zou 외 arxiv

We introduce TerminalWorld, a scalable data engine that automatically reverse-engineers high-fidelity evaluation tasks from "in-the-wild" terminal recordings. Processing 80,870 terminal recordings, the engine yields a fu…

MMTB: Evaluating Terminal Agents on Multimedia-File Tasks

2026-05-08 · Chiyeong Heo, Jaechang Kim, Junhyuk Kwon, Hoyoung Kim 외 arxiv

Terminals provide a powerful interface for AI agents by exposing diverse tools for automating complex workflows, yet existing terminal-agent benchmarks largely focus on tasks grounded in text, code, and structured files.…

A Self-Evolving Framework for Efficient Terminal Agents via Observational Context Compression

2026-04-21 · Jincheng Ren, Siwei Wu, Yizhi Li, Kang Zhu 외 arxiv

As terminal agents scale to long-horizon, multi-turn workflows, a key bottleneck is not merely limited context length, but the accumulation of noisy terminal observations in the interaction history. Retaining raw observa…

SETA: Scaling Environments for Terminal Agents

2026-07-12 · Qijia Shen, Zhiqi Huang, Vamsidhar Kamanuru, Aznaur Aliev 외 arxiv

Large language models (LLMs) are rapidly shifting toward agents that solve tasks through diverse interfaces, including web and graphical user interfaces (GUIs). Among these, the terminal command line provides a text-base…

Reinforcement Learning

Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading

2026-07-09 · Zongxia Li, Zhongzhi Li, Yucheng Shi, Ruhan Wang 외 arxiv

AI agents have become capable of autonomously completing short, well-specified tasks. However, existing terminal benchmarks largely focus on simple problems that finish within minutes and are evaluated only by their fina…