Benchmark RadarAI BENCHMARK PROFILE
FinED-Bench
FinED-Bench evaluates error detection in financial documents across nine real-world financial scenarios, covering three cognitive complexity levels. It includes over 900 documents from 2025, with data and code publicly available.
- Released
- 2026-06-03
- Readiness
- Runnable
- Primary field
- Finance & Economics
Why it matters
Financial document accuracy is critical for compliance and decision-making, yet current LLMs struggle with error detection. FinED-Bench provides a fixed protocol to assess and improve model reliability in this domain.
Motivation
Ensuring the accuracy of financial documents is critical for economic analysis, regulatory compliance, and corporate decision-making.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.