OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Chengyu Shen, Yujie Fu, Gangtao Xin, Yanheng Hou, Wenlong Fei, Guojie Zhu, Jiawei Li, Hongcheng Gao, Runming He, Zhen Hao Wong, Meiyi Qiang, Hao Liang, Zhao Cao, Hao Jiang, Chong Chen, Wentao Zhang
• Venue: arXiv: cs.CL
• Published: 2026-07-16
• Collected: 2026-07-17
• URL: http://arxiv.org/abs/2607.14989v1
• PDF: https://arxiv.org/pdf/2607.14989v1
• DOI: UNKNOWN

Plain-Language Brief

Researchers have introduced OmniaBench, a new benchmark for evaluating the abilities of general AI agents across various scenarios with explicit state spaces. The benchmark includes a hierarchical taxonomy covering ToC, ToB, and ToE domains, and it consists of 1431 tasks. The results show that even advanced models like Claude-Sonnet-5 and GPT-5.6-Sol struggle with the benchmark, achieving Overall Pass@1 scores of only 58.54 and 57.14 respectively.

What the Document Says

• The text introduces OmniaBench, a benchmark for evaluating general AI agents.
• OmniaBench covers diverse scenarios with explicit state spaces.
• Application-oriented scenario knowledge is derived from various sources.
• A hierarchical taxonomy of 90 level-1 and 354 level-2 domains spans ToC, ToB, and ToE.
• Executable environments are constructed and tasks are synthesized through four routes: DAG, DAG-S, Solver, and Program.
• OmniaBench includes a ten-dimensional capability taxonomy and eight compositional atomic difficulty factors for fine-grained evaluation.
• The dataset contains 1431 tasks, with a challenging subset of 644 tasks.
• Two specific models, Claude-Sonnet-5 and GPT-5.6-Sol, were tested and their Overall Pass@1 scores are given.
• Further analyses reveal differences across domains, capabilities, planning limitations, constraint maintenance issues, and adaptive correction challenges.

Key Concepts Mentioned

• OmniaBench
• General AI agents
• Hierarchical taxonomy
• Executable environments
• Scenario knowledge
• DAG
• DAG-S
• Solver
• Program
• Capability taxonomy
• Atomic difficulty factors
• Task dataset
• Claude-Sonnet-5
• GPT-5.6-Sol
• Overall Pass@1 scores
• Analyses
• Domains
• Capabilities
• Planning
• Constraint maintenance
• Adaptive correction

Claims Made by the Document

• OmniaBench provides a benchmark for evaluating general AI agents.
• The benchmark covers diverse scenarios with explicit state spaces.

Evidence Used

• 'We introduce OmniaBench, a benchmark for evaluating general agents across diverse scenarios with explicit state spaces.'

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Large language models are increasingly evolving from text generators into general agents capable of understanding user requests, invoking external tools, and completing complex tasks through interaction. However, existing agent benchmarks often focus on limited scenarios, tool ecosystems, or interaction formats, making it difficult to systematically characterize model capabilities across heterogeneous application settings. We introduce OmniaBench, a benchmark for evaluating general agents across diverse scenarios with explicit state spaces. We derive application-oriented scenario knowledge from app stores, product documents, industry resources, Web retrieval, and human refinement, forming a hierarchical taxonomy that spans ToC, ToB and ToE with 90 level-1 and 354 level-2 domains. Based on this taxonomy, we construct executable environments and synthesize single-turn and multi-turn tasks through four complementary routes: DAG, DAG-S, Solver, and Program. OmniaBench further introduces a ten-dimensional capability taxonomy and eight compositional atomic difficulty factors to support fine-grained evaluation and analysis. The resulting dataset contains 1,431 tasks, together with a challenging subset of 644 tasks designed to reduce evaluation cost and mitigate potential contamination of the full set after public release. The bench presents substantial challenges to current frontier models, with even Claude-Sonnet-5 and GPT-5.6-Sol achieving Overall Pass@1 scores of only 58.54 and 57.14, respectively. Further analyses reveal clear differences across domains and capabilities, as well as persistent limitations in planning, constraint maintenance, and adaptive correction. OmniaBench provides a broad and diagnostic benchmark for characterizing the capability boundaries of general agents.