TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks

• URL: http://arxiv.org/abs/2605.22535v1
• Date: Thu, 21 May 2026 14:24:43 GMT
• Title: TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks
• Abstract summary: TerminalWorld is a scalable data engine that automatically reverse-engineers high-fidelity evaluation tasks from "in-the-wild" terminal recordings.<n>The engine yields a full benchmark of 1,530 validated tasks, spanning 18 real-world categories.<n>TerminalWorld captures real-world terminal capabilities distinct from existing expert-Verified benchmarks.
• Score: 23.863417507169697
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: 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 full benchmark of 1,530 validated tasks, spanning 18 real-world categories, ranging from short everyday operations to workflows exceeding 50 steps, and covering 1,280 unique commands. From these, we curate a Verified subset of 200 representative, manually reviewed tasks. Comprehensive benchmarking on TerminalWorld-Verified across eight frontier models and six agents reveals that current systems still struggle with authentic terminal workflows, achieving a maximum pass rate of only 62.5%. Moreover, TerminalWorld captures real-world terminal capabilities distinct from existing expert-curated benchmarks (e.g., Terminal-Bench), with only a weak correlation to their scores (Pearson r=0.20). The automated engine makes TerminalWorld authentic and scalable by construction, enabling it to evaluate agents in real-world terminal environments as developer practices evolve. Data and code are available at https://github.com/EuniAI/TerminalWorld.

Related papers

• OSWorld2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks [98.07146747892239]
  OSWorld 2.0 is a benchmark of 108 long-horizon computer-use across everyday and professional tasks.<n>Each task represents a realistic end-to-end workflow that takes human users a median of about 1.6 hours to complete.
  arXiv  Detail & Related papers  (2026-06-28T17:59:17Z)
• TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents [37.00898251261264]
  We introduce TUA-Bench, a general-purpose benchmark for terminal-use agents.<n> TUA-Bench includes 120 real-world tasks across five task families.<n>We find that the strongest frontier agent achieves 65.8% overall performance.
  arXiv  Detail & Related papers  (2026-06-26T17:59:51Z)
• Mind the Gap: Can Frontier LLMs Pass a Standardized Office Proficiency Exam? [90.39633603556767]
  Large Language Model (LLM) agents for computer automation is accelerating, yet their ability to navigate complex, professional-grade productivity software is largely untested.<n>To quantify this capability, we introduce an evaluation based on China's National Computer Rank Examination (NCRE)<n>We benchmark 7 frontier LLMs and observe stark limitations: single-turn models score a maximum of 36.6%.<n>A stronger agentic system with execution feedback, iterative repair, and broader Office automation access reaches 68.8%, but remains below the 95.5% community-reference score used as a scoring sanity check.
  arXiv  Detail & Related papers  (2026-06-09T14:59:14Z)
• Terminal-World: Scaling Terminal-Agent Environments via Agent Skills [52.39713754337834]
  We introduce Terminal-World, a fully automated pipeline that uses agent skills as the central synthesis primitive.<n>We construct 5,723 training environments and train Terminal-World-8B/14B/32B, evaluated across 6 benchmarks.<n>Terminal-World-32B surpasses Nemotron-Terminal-32B on Terminal-Bench 2.0 by +4.5 Pass@1 (31.5) and achieves 43.8 Pass@3.
  arXiv  Detail & Related papers  (2026-05-20T08:14:51Z)
• WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation [88.10947115397971]
  This work presents WildClawBench, a native-runtime benchmark of 60 human-authored, bilingual, multimodal tasks spanning six thematic categories.<n>Each task averages roughly 8 minutes of wall-clock time and over 20 tool calls.<n> Grading is hybrid, combining deterministic rule-based checks, environment-state auditing of side effects, and an LLM/VLM judge for semantic verification.
  arXiv  Detail & Related papers  (2026-05-11T17:49:43Z)
• Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces [126.23612941699565]
  Terminal-Bench 2.0 is a benchmark composed of 89 tasks in computer terminal environments inspired by problems from real world.<n>We show that frontier models and agents score less than 65% on the benchmark.<n>We publish the dataset and evaluation harness to assist developers and researchers in future work at https://www.tbench.ai/.
  arXiv  Detail & Related papers  (2026-01-17T01:29:30Z)
• Step-GUI Technical Report [84.83795946544292]
  We introduce a self-evolving training pipeline powered by the Calibrated Step Reward System.<n>We also introduce Step-GUI, a family of models that achieves state-of-the-art GUI performance.<n>To assess whether agents can handle authentic everyday usage, we introduce AndroidDaily.
  arXiv  Detail & Related papers  (2025-12-17T13:26:30Z)
• CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents [49.68117560675367]
  Crab is the first benchmark framework designed to support cross-environment tasks.<n>Our framework supports multiple devices and can be easily extended to any environment with a Python interface.<n>The experimental results demonstrate that the single agent with GPT-4o achieves the best completion ratio of 38.01%.
  arXiv  Detail & Related papers  (2024-07-01T17:55:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.