Agentic benchmark
QwenClawBench leaderboard
Every model the catalog carries a published QwenClawBench value for, ranked by that value.
CategoryAgentic
MeasureEnd-to-end agent evaluation
TasksReal-world agent workflows
DifficultyBroad real-world agentic execution
Published byQwen3.6 launch benchmarks
Qwen's internal OpenClaw-style benchmark for measuring broad real-world agent performance across practical productivity and research tasks.
QwenClawBench ranking
10 models with a published QwenClawBench value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | End-to-end agent evaluation |
|---|---|---|---|
| 1 |  Qwen3.7 Max | Alibaba | 64.3 |
| 2 | Qwen3.7 Plus | Alibaba | 61.8 |
| 3 | Qwen 3.6 Max (preview) | Alibaba | 59.0 |
| 4 | Qwen3.6 Plus | Alibaba | 57.2 |
| 5 |  Kimi K2.5 | Moonshot AI | 54.3 |
| 6 |  GLM-5 | Z.AI | 54.1 |
| 7 | Qwen3.6-27B | Alibaba | 53.4 |
| 8 | Qwen3.6-35B-A3B | Alibaba | 52.6 |
| 9 |  Claude Opus 4.5 | Anthropic | 52.3 |
| 10 | Qwen3.5 397B | Alibaba | 51.8 |
Evidence key: Observed
Rows are ordered by the value Qwen3.6 launch benchmarks published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.