spreadsheetbench

The spreadsheetbench tag covers discussions about the SpreadsheetBench benchmark, which evaluates AI models on real-world spreadsheet editing tasks. Content under this tag includes Google's announcement that its Gemini model achieved a 70.48% pass rate on SpreadsheetBench, demonstrating state-of-the-art proficiency in creating, organizing, and editing spreadsheets from plain-English instructions. This tag is relevant for users interested in AI performance benchmarks, spreadsheet automation, and the capabilities of large language models in practical office tasks.