Agent Performance on OfficeQA Pro V2Cost versus correctness for model + harness combinations on OfficeQA Pro V2. Scaling OfficeQA Pro V2 with Synthetic DataSynthetic Data Pipeline used to build OfficeQA Pro V2. OfficeQA Pro refers to the frontier split of OfficeQA questions, OfficeQA Full is a combined set including easier questions, and OfficeQA Pro V2 is our newly released benchmark. Although built on a new corpus, OfficeQA Pro V2 preserves the core enterprise grounded-reasoning capabilities measured by the original OfficeQA benchmark, both across the full question set, OfficeQA Full, and within its frontier-difficulty subset, OfficeQA Pro. In total, 74.4% of OfficeQA Pro V2 questions require four or more sources, compared with 62.4% in OfficeQA Pro and 56.1% in OfficeQA Full.