Blog
Insights on AI in investment banking, benchmark updates, and model performance analysis.
Claude Opus 4.5 vs ChatGPT 5.2: Performance on Hard Valuation Tasks
A deep dive comparing how the latest frontier models handle complex LBO modeling, DCF analysis, and financial statement parsing in IB-Bench.
Why We Weight Hard Tasks 45%: The Case for Difficulty-Based Scoring
Explaining the methodology behind IB-Bench's scoring system and why complex tasks deserve more weight than simple calculations.
Introducing IB-Bench: A Real-World Benchmark for AI in Investment Banking
Why generic benchmarks fail to capture what matters in finance, and how IB-Bench measures the tasks that actually define analyst work.