Benchmark
CoWorkBench
What is CoWorkBench?
CoWorkBench is Qwen's internal cowork benchmark for evaluating long-horizon office and productivity agent tasks across domains such as computer science…
- Modality
- text
- Evaluates
- productivity, reasoning, agents
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 4
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/coworkbench?top_n=500
CoWorkBench results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen3.7 Max | Qwen | 0.672 | 2026-05-19 | evidence |
| Qwen3.7-Plus | Qwen | 0.651 | 2026-05-31 | evidence |
| Qwen3.8 Max | Qwen | 0.748 | 2026-08-02 | evidence |
| Qwen3.8-27B | Qwen | 0.707 | 2026-08-14 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.