Cua Launches Driver Bench for Background Desktop Agents

Cua, together with Scale AI's AgentEnv, released Driver Bench, a benchmark testing whether agents can complete desktop tasks in the background without disturbing users. On macOS, Cua achieved a 58% pass rate, far ahead of Codex at 33%.

2026-11-07 ~ 2026-11-07 · 2 related posts

• Cua Driver Bench: 58% pass rate on background macOS desktop tasks vs Codex 33% — TianbaoX · 2026-11-07

1 near-duplicate retellings: victor_explore