# OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks

arXiv: https://arxiv.org/abs/2606.29537 · HTML: https://mlsanity.com/paper/2606.29537/

## Key facts

- Cited by 40 papers in the map (data through 14 Oct 2026).
- Placed in GUI and mobile agents, one of 16 directions in LLM agents and tool use.
- Builds on 77 of its 102 references that are in the map.
- First posted on arXiv 28 Jun 2026.

## Placed in

- LLM agents and tool use / [GUI and mobile agents](https://mlsanity.com/agents/gui-and-mobile-agents/)

## Builds on (most cited first, 30 of 77)

- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs](https://mlsanity.com/paper/2307.16789/) (2023)
- [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024)
- [AgentBench: Evaluating LLMs as Agents](https://mlsanity.com/paper/2308.03688/) (2023)
- [WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents](https://mlsanity.com/paper/2207.01206/) (2022)
- [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [GAIA: a benchmark for General AI Assistants](https://mlsanity.com/paper/2311.12983/) (2023)
- [CogAgent: A Visual Language Model for GUI Agents](https://mlsanity.com/paper/2312.08914/) (2023)
- [BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agents](https://mlsanity.com/paper/2504.12516/) (2025)
- [τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://mlsanity.com/paper/2506.07982/) (2025)
- [UI-TARS: Pioneering Automated GUI Interaction with Native Agents](https://mlsanity.com/paper/2501.12326/) (2025)
- [Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces](https://mlsanity.com/paper/2601.11868/) (2026)
- [SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents](https://mlsanity.com/paper/2401.10935/) (2024)
- [GPT-4V(ision) is a Generalist Web Agent, if Grounded](https://mlsanity.com/paper/2401.01614/) (2024)
- [MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering](https://mlsanity.com/paper/2410.07095/) (2024)
- [AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agents](https://mlsanity.com/paper/2405.14573/) (2024)
- [AppAgent: Multimodal Agents as Smartphone Users](https://mlsanity.com/paper/2312.13771/) (2023)
- [AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents](https://mlsanity.com/paper/2407.18901/) (2024)
- [OS-ATLAS: A Foundation Action Model for Generalist GUI Agents](https://mlsanity.com/paper/2410.23218/) (2024)
- [Language Models can Solve Computer Tasks](https://mlsanity.com/paper/2303.17491/) (2023)
- [MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation](https://mlsanity.com/paper/2310.03302/) (2023)
- [ScienceWorld: Is your Agent Smarter than a 5th Grader?](https://mlsanity.com/paper/2203.07540/) (2022)
- [TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks](https://mlsanity.com/paper/2412.14161/) (2024)
- [Android in the Wild: A Large-Scale Dataset for Android Device Control](https://mlsanity.com/paper/2307.10088/) (2023)
- [Agent-SafetyBench: Evaluating the Safety of LLM Agents](https://mlsanity.com/paper/2412.14470/) (2024)
- [SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?](https://mlsanity.com/paper/2509.16941/) (2025)
- [Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction](https://mlsanity.com/paper/2412.04454/) (2024)
- [ScreenSpot-Pro: GUI Grounding for Professional High-Resolution Computer Use](https://mlsanity.com/paper/2504.07981/) (2025)
- [GUI-R1 : A Generalist R1-Style Vision-Language Action Model For GUI Agents](https://mlsanity.com/paper/2504.10458/) (2025)

## Built on by (most cited first, 30 of 40)

- [Kimi K3: Open Frontier Intelligence](https://mlsanity.com/paper/2607.24653/) (2026)
- [Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents](https://mlsanity.com/paper/2607.28227/) (2026)
- [MerchantBench: Benchmarking LLM Agents for Long-Term Coherence in E-Commerce Operations](https://mlsanity.com/paper/2607.28956/) (2026)
- [StateAct: Program State, before Pixels, for Long-Horizon Computer-Use Agents](https://mlsanity.com/paper/2607.22798/) (2026)
- [The Physics of Multi-Turn Long-Horizon Planning: From Pre-training to Post-training via Single- and Multi-Teacher On-Policy Agentic Distillation](https://mlsanity.com/paper/2607.24720/) (2026)
- [UI-Venus-2 Technical Report](https://mlsanity.com/paper/2609.00028/) (2026)
- [CUA-Universe: A Scalable and Dynamic Environment for Hybrid GUI+CLI Agents](https://mlsanity.com/paper/2609.05374/) (2026)
- [SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents](https://mlsanity.com/paper/2607.23263/) (2026)
- [UI-Mate: Advancing Open-Weight Foundation GUI Agents with In-Context Demonstrations](https://mlsanity.com/paper/2608.15930/) (2026)
- [CADWorld: Computer-Use Benchmark for Long-Horizon Computer-Aided Design](https://mlsanity.com/paper/2609.16251/) (2026)
- [EchoPath: Execution-Level Replayable Memory for GUI Agents](https://mlsanity.com/paper/2609.16635/) (2026)
- [An Exam for Active Observers](https://mlsanity.com/paper/2607.16165/) (2026)
- [Desktop-Delta Bench: Do Computer-Use Models Understand Desktop GUI Transitions?](https://mlsanity.com/paper/2607.26041/) (2026)
- [OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution](https://mlsanity.com/paper/2608.00677/) (2026)
- [Spine-Branch Coordination for Multi-agent Computer Use](https://mlsanity.com/paper/2608.22077/) (2026)
- [Benchmarking General Mobile Assistants in Challenging Real-World Scenarios](https://mlsanity.com/paper/2608.27477/) (2026)
- [Solaris: Towards Interfaces That Are Generated, Not Coded](https://mlsanity.com/paper/2609.00776/) (2026)
- [Beyond Task Completion: Training Capable and Safe Computer-Use Agents](https://mlsanity.com/paper/2609.22178/) (2026)
- [The GUI Is Not the State: Diagnosing State Aliasing in GUI World Models](https://mlsanity.com/paper/2609.32679/) (2026)
- [CUA-Sandbox: Efficient Environments for Computer-Use Agent Reinforcement Learning](https://mlsanity.com/paper/2609.32750/) (2026)
- [On the Behavioral Traits of LLM Agents](https://mlsanity.com/paper/2609.32776/) (2026)
- [Does Learning to Predict the World Help Agents Act? Auditing World-Model Post-Training](https://mlsanity.com/paper/2609.33335/) (2026)
- [Beyond Skill Evolution: Self-Evolving Context Management Policies for Long-Horizon Agent Harnesses](https://mlsanity.com/paper/2609.34649/) (2026)
- [APEX-Voice: Can Voice Agents Complete Professional Workflows Through Full-Duplex Interaction](https://mlsanity.com/paper/2609.34973/) (2026)
- [SCOUT: Synergizing Reasoning and Tool-Use for Computer-Use Safety](https://mlsanity.com/paper/2609.36201/) (2026)
- [Neuro-Symbolic Computer Use: Learning Reusable Policies for Reliable and Efficient Execution](https://mlsanity.com/paper/2609.36927/) (2026)
- [EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments?](https://mlsanity.com/paper/2609.37686/) (2026)
- [OSWorld-Science: A Benchmark of Computer Use Agents for Learning and Using Scientific Software](https://mlsanity.com/paper/2609.39903/) (2026)
- [cua-speedrun: Standardized Benchmarking of the Speed of Computer-Use Agents](https://mlsanity.com/paper/2609.40284/) (2026)
- [World Editing: Intervening on Executable Worlds at Increasing Depth](https://mlsanity.com/paper/2610.02331/) (2026)

## Abstract (arXiv)

Existing computer-use benchmarks fail to capture the realism, complexity, and long-horizon demands of real-world computer use, limiting their ability to reveal the limitations of frontier agents. We introduce OSWorld 2.0, a benchmark of 108 long-horizon computer-use workflows across everyday and professional tasks, designed to capture complex and challenging real-world phenomena. Each task represents a realistic end-to-end workflow that takes human users a median of about 1.6 hours to complete and requires an average of 318 tool calls with Claude Opus 4.7 using maximum thinking, compared with about 30 in OSWorld 1.0. OSWorld 2.0 targets challenge phenomena that are common in real workflows yet underrepresented in prior benchmarks, spanning interaction-design challenges such as streaming interaction and dynamic environments, as well as agent-pattern challenges such as cross-source reasoning, implicit-state inference, and visual-spatial precision. Tasks are grounded in authentic input artifacts and cross-referenced against realistic stateful user profile data, and include separate safety reports auditing safety-sensitive execution. Under our primary binary-completion metric at 500 steps, Claude Opus 4.8 with maximum thinking and batched tool calls scores best but still completes only 20.6% of tasks at a 54.8% partial score; GPT-5.5 is far more token-efficient yet plateaus near 13%. These results show that current agents are still far from professional-level computer use: rather than stumbling on basic GUI control or coding, they lose track of constraints, miss information that arrives mid-task, guess rather than ask the user, and skip verification, struggling most when a task hinges on hidden state they must recover.