Harness Decides AI Agent Performance! Wall Street Tests Qwen Office and Workbuddy

Harness advantages are enough to offset the gap in model intelligence. Taking first place was Alibaba's Qwen Office, despite its underlying model intelligence score ranking only fourth; Claude Opus 5, which topped model capability, saw its Agent product Claude Cowork take second place. Another counterintuitive result is that Workbuddy, recently gaining much attention, had the lowest implied Harness score among Chinese Agents in this round of testing.