Finance and analyst work
The research an analyst does: reading company filings, finding figures and answering questions with them. Good means the answer is correct and supported by the filing.
Updated every Friday. Sources last checked 2026-09-26.
At a glance
Vendor claims have not been collected for any category yet.
ⓘ More about this category
- Reliability measured by a linked source; figures not shown here
- 1 source, linked results only
- Sources last checked 2026-09-24
- Ring 4: needs rings 1 to 3 first
- VFA = Finance Agent v2
Who does well
Finance Agent v2 publishes results for this. We don't have permission to show its figures here. See their results below
Findings for this category aren't written yet.
Tested by
Their figures are not shown here because permission to reuse them is not yet confirmed.
Finance Agent v2
Vals AI, checked 2026-09-24A severity-weighted average of checks on answers to financial analyst questions. Missing a dealbreaker check gives the whole question zero credit; other missed checks reduce the credit.
Vals AI reports stronger results on straightforward filing lookups than on multi-step financial analysis that needs exact figures and industry judgment.
Last updated 2026-09-27
See their resultsMore about this test: Finance Agent v2
Reliability: Vals runs each model three times on the test set and reports the mean score with its standard error.
Technical details for Finance Agent v2
Accuracy (Partial Credit): A severity-weighted average of checks on answers to financial analyst questions. Missing a dealbreaker check gives the whole question zero credit; other missed checks reduce the credit.
License: Score reuse not stated. Checked 2026-09-24.
What this doesn't tell you
- One source covers this task so far.
- None of these tests used your filings or your firm's methods.
- Scores from different sources are not comparable, even when they share a unit.
- None of these tests were run on your own documents, code or data.
- No source for this task allows us to show its figures yet. Follow the links for their results.
Data version 2026-09-30+832fb354fcbe · Terms of use