As large language models evolve into general agents capable of understanding user requests, invoking external tools, and completing complex tasks through interaction, 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. Hence, 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, along with a challenging subset of 644 tasks designed to reduce evaluation cost and mitigate potential contamination of the full set after public release.
The benchmark 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.
Blogger's Review: OmniaBench fills a significant gap in existing evaluation tools, particularly in assessing capabilities in diverse scenarios. This research not only offers a reference framework for developing more powerful AI agents but also provides valuable data support for future studies. Its challenging task set will drive advancements in AI models' handling of complex tasks.