Single-step operation description prediction on GUIOdyssey
78.22AATeach VLM
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Teach VLM2026.06 | 78.22 | 45.55 | 78.28 | 42.99 | |
| Doubao-Seed-1.82026.06 | 67.15 | 32.07 | 34.32 | 13.97 | |
| Doubao-Seed-2.02026.06 | 66.63 | 31.25 | 43.43 | 16.55 | |
| Qwen3-VL-8B2026.06 | 34.61 | 36.42 | 30.03 | 9.57 | |
| Qwen3.5-9B2026.06 | 32.59 | 40.96 | 26.14 | 11.9 | |
| GPT-4o2026.06 | 11.48 | 40.14 | 3.35 | 2.12 |