Benchmark
DAYJOB: Tax
Evaluate agents on professional tax assignments with complex records and incomplete instructions.
Explore the original evidence
Explore the original evaluation, its methods, and reported findings.
Open evalCoverage
Fifty environments span eight areas of tax practice, from individual and business filings to international tax and disputes. This is the Tax member of the DAYJOB family, separate from Finance and Healthcare.
Methodology
Tax experts build and review assignments and grading rubrics. Agents investigate records and produce filing packages, workpapers, reviews or advice. Reported pass@1 requires at least 95% of task criteria.