Finance Agent: Results as Published

Finance Agent tests AI agents on research tasks that a financial analyst would do, such as answering questions from recent SEC filings. The original version had 537 expert-written questions in nine task categories, and the agent is given search and EDGAR database access. Version 2 contains 927 expert-reviewed questions, and results are reported as partial credit or as all-pass, where a fully correct answer is required.

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.