Scene-level script generation on V2S benchmark 5-minute videos 1.0 (test)
58.8Localization ScoreGemini-3-pro
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Gemini-3-proOmni=✓2026.04 | 58.8 | 63.1 | 46.9 | 61.6 | 54.8 | 57 | 75.3 | |
| Seed-1.8Omni=×2026.04 | 57.9 | 58.6 | 47.7 | 58.7 | 52.8 | 55.1 | 74 | |
| Seed-2.0-proOmni=×2026.04 | 57.7 | 62.2 | 49.2 | 62.7 | 54.3 | 57.2 | 75.5 | |
| Gemini-2.5-proOmni=✓2026.04 | 56.6 | 62.4 | 50.8 | 60.1 | 54.6 | 56.9 | 74.1 | |
| Gemini-3-flashOmni=✓2026.04 | 54.6 | 59.8 | 42.7 | 54.9 | 50.4 | 52.5 | 70.3 | |
| OmniScriptParam./B=8, Omni=✓2026.04 | 54 | 58.4 | 41.9 | 58.1 | 49.5 | 52.4 | 74.6 | |
| Gemini-2.5-flashOmni=✓2026.04 | 52.8 | 57.1 | 45.7 | 56.1 | 50.3 | 52.3 | 69.6 | |
| Qwen3VLParam./B=235/22, Omni=×2026.04 | 52.6 | 60.2 | 45.4 | 57.9 | 50.9 | 53.4 | 72.8 | |
| Qwen3VLParam./B=32, Omni=×2026.04 | 50.4 | 58.7 | 42.7 | 55.4 | 47.9 | 51 | 71.1 | |
| TimeChat-CaptionerParam./B=8, Omni=✓, Segmented protocol (†)=yes2026.04 | 47.3 | 55.5 | 41.8 | 47.1 | 48.4 | 48 | 69.6 | |
| Qwen3VLParam./B=8, Omni=×2026.04 | 41.3 | 49.7 | 31.8 | 39.8 | 41.7 | 40.9 | 60.6 | |
| MiniCPM-O-4.5Param./B=9, Omni=✓, Segmented protocol (†)=yes2026.04 | 39.9 | 51.7 | 34.7 | 45.2 | 43.8 | 43.1 | 63.4 | |
| TimeChat-CaptionerParam./B=8, Omni=✓, Segmented protocol (†)=no2026.04 | 19.9 | 29.5 | 17.3 | 28.8 | 30.5 | 25.2 | 46.6 | |
| Qwen3-OmniParam./B=30/3, Omni=✓2026.04 | 18.4 | 26 | 14.6 | 23.6 | 22.4 | 21 | 29.6 | |
| MiniCPM-O-4.5Param./B=9, Omni=✓, Segmented protocol (†)=no2026.04 | 10.3 | 22 | 8.4 | 17.7 | 17.4 | 15.1 | 32 |