Benchmark RadarAI BENCHMARK PROFILE
Tencent WorkBuddy Bench
Multi-domain coding-agent benchmark with reverse-engineered tasks across Code, Web, Office, and Security. Each subset has its own scoring instrument; scores are not comparable across subsets.
- Released
- 2026-07-23
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
Addresses contamination in coding-agent evaluation by constructing tasks not discoverable via web search, and provides a reproducible open-source framework for auditable third-party runs.
Motivation
We introduce Tencent WorkBuddy Bench, a multi-domain evaluation suite for coding agents; this report documents its construction methodology, scoring protocol, and a cross-model leaderboard.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.