WorkBuddy-Bench
A benchmark for measuring agent performance on real office tasks. The tasks cover multi-step tool-calling scenarios such as retrieval, spreadsheet processing, document generation, and cross-application operations.
Timeline (last 90 days)
Google Research reported a performance improvement of up to +11.5 points on the WorkBuddy-Bench benchmark.
According to official disclosures, a single refinement step alone improved GLM-5 by 11.5 points on WorkBuddy-Bench.
This 11.5-point improvement result comes only from a single source's official disclosure and currently lacks third-party reproduction.
All Facts (3)
Google Research reported a performance improvement of up to +11.5 points on the WorkBuddy-Bench benchmark.
50%UnverifiedAccording to official disclosures, a single refinement step alone improved GLM-5 by 11.5 points on WorkBuddy-Bench.
50%UnverifiedThis 11.5-point improvement result comes only from a single source's official disclosure and currently lacks third-party reproduction.
50%