Dynamic Computer Use on DynaCU-Bench 100-task
22Standard AccuracyQwen3-VL-235B-A22B-Instruct
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen3-VL-235B-A22B-InstructModel=Qwen3-VL-235B-A22B-Instruct, Provider=DeepInfra, Access=OpenRouter2026.06 | 22 | 64 | 42 | 4 | 46 | 4.5 | |
| Qwen3-VL-30B-A3B-InstructModel=Qwen3-VL-30B-A3B-Instruct, Provider=DeepInfra, Access=OpenRouter2026.06 | 18 | 42 | 24 | 8 | 32 | 1.8 |