Query-to-Presentation Generation on PresentEval Single Presentation 1.0 (test)
4.84Quiz ScorePresentAgent-2
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PresentAgent-2Model=Qwen3.5-VL-Plus2026.05 | 4.84 | 4.5 | 4.48 | 4.43 | 4.47 | |
| PresentAgent-2Model=GPT-5.52026.05 | 4.83 | 4.25 | 4.3 | 4.2 | 4.25 | |
| Human ReferenceModel=Human-created2026.05 | 4.82 | 4.54 | 4.45 | 4.4 | 4.46 | |
| PresentAgent-2Model=Claude Opus 4.72026.05 | 4.8 | 4.47 | 4.47 | 4.35 | 4.43 | |
| PresentAgent-2Model=Gemini 3.1 Pro2026.05 | 4.78 | 4.45 | 4.24 | 4.37 | 4.35 | |
| PresentAgent-2Model=GLM-4.7V2026.05 | 4.75 | 4.21 | 4.21 | 4.13 | 4.18 |