Someone benchmarked on how accurate different AI are on excel documents
A new benchmark called SpreadsheetBench evaluates AI models on their accuracy in handling Excel documents. The benchmark uses real-world tasks from Excel forums, requiring exact cell-by-cell accuracy and testing complex formula dependencies and structural reorganization. Specialized AI tools like Dealglass and Leni achieved over 90% accuracy, significantly outperforming general models such as Claude Opus 4.6 (around 80%) and GPT 5.4 (high 70s). AI
IMPACT Specialized AI tools demonstrate superior performance in complex spreadsheet tasks, suggesting a need for domain-specific solutions over general models for business applications.