Desktop Cleaning on Real-world robot tasks 1.0 (test)
8Success RateGPT-4o
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-4oTool Augmentation=w/2026.05 | 8 | |
| GPT-5Tool Augmentation=w/2026.05 | 7 | |
| Claude-3.7Tool Augmentation=w/2026.05 | 6 | |
| Claude-4.6Tool Augmentation=w/2026.05 | 6 | |
| Qwen3-8BTool Augmentation=w/2026.05 | 6 | |
| Qwen3-32BTool Augmentation=w/2026.05 | 6 | |
| Qwen2.5-32BTool Augmentation=w/2026.05 | 5 | |
| Qwen3.5-35BTool Augmentation=w/2026.05 | 4 | |
| GPT-4oTool Augmentation=w/o2026.05 | 0 | |
| GPT-5Tool Augmentation=w/o2026.05 | 0 | |
| Claude-3.7Tool Augmentation=w/o2026.05 | 0 | |
| Claude-4.6Tool Augmentation=w/o2026.05 | 0 | |
| Qwen3-8BTool Augmentation=w/o2026.05 | 0 | |
| Qwen3-32BTool Augmentation=w/o2026.05 | 0 | |
| Qwen2.5-32BTool Augmentation=w/o2026.05 | 0 | |
| Qwen3.5-35BTool Augmentation=w/o2026.05 | 0 |