Audiovisual Video Captioning on UGC-VideoCap
80.4Audio ScoreGemini-3-Pro
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gemini-3-ProModality=A + V2026.02 | 80.4 | 84.7 | 80.6 | 81.9 | |
| ASID-CaptionerSize=7B, Modality=A + V2026.02 | 79.1 | 84.4 | 80.2 | 81.2 | |
| ASID-CaptionerSize=3B, Modality=A + V2026.02 | 78.6 | 84.8 | 80.2 | 81.2 | |
| DiaDemSize=7B2026.01 | 75.4 | 76.8 | 74.6 | 75.6 | |
| AVoCaDOSize=7B2026.01 | 73 | 74.6 | 71.8 | 73.2 | |
| AVoCaDOSize=7B, Modality=A + V2026.02 | 73 | 74.6 | 71.8 | 73.2 | |
| OmniCaptioner-IFParameters=7B2026.06 | 69.79 | 75.94 | 73.19 | 72.97 | |
| Gemini-3-ProSize=-2026.01 | 69.7 | 75.6 | 74.3 | 73.2 | |
| Gemini-2.5-ProSize=-2026.01 | 69.5 | 74.7 | 73.7 | 72.6 | |
| Gemini-2.5-ProModality=A + V2026.02 | 69.5 | 74.7 | 73.7 | 72.6 | |
| Gemini-2.5-Pro2026.06 | 69.5 | 74.7 | 73.7 | 72.6 | |
| Gemini-2.5-FlashSize=-2026.01 | 69.1 | 75.8 | 74 | 73 | |
| Gemini-2.5-FlashModality=A + V2026.02 | 69.1 | 75.8 | 74 | 73 | |
| Qwen3-Omni-CaptionerSize=30B-A3B2026.01 | 69 | 75.5 | 72.3 | 72.5 | |
| Qwen3-Omni-CaptionerSize=30B-A3B, Modality=A + V2026.02 | 69 | 75.5 | 72.3 | 72.3 | |
| OmniCaptioner-IFParameters=3B2026.06 | 67.71 | 73.91 | 70.43 | 70.68 | |
| Qwen3-Omni-InstructSize=30B-A3B2026.01 | 67.5 | 74.8 | 72.3 | 71.5 | |
| Qwen3-Omni-InstructSize=30B-A3B, Modality=A + V2026.02 | 67.5 | 74.8 | 72.3 | 71.5 | |
| Qwen3-OmniParameters=30B, Active Parameters=A3B, Mode=Instruct2026.06 | 67.5 | 74.8 | 72.3 | 71.5 | |
| video-SALMONN-2Size=7B2026.01 | 61.8 | 71.4 | 68.5 | 67.2 | |
| video-SALMONN-2Size=7B, Modality=A + V2026.02 | 61.8 | 71.4 | 68.5 | 67.2 | |
| video-SALMONN-2Parameters=7B2026.06 | 61.8 | 71.4 | 68.5 | 67.2 | |
| UGC-VideoCaptionerSize=3B2026.01 | 61.4 | 58.4 | 57.5 | 59.1 | |
| UGC-VideoCaptionerSize=3B, Modality=A + V2026.02 | 61.4 | 58.4 | 57.5 | 59.1 | |
| ARC-Hunyuan-VideoSize=7B, Modality=A + V2026.02 | 52.7 | 56 | 55.8 | 54.8 | |
| Qwen2.5-OmniParameters=3B2026.06 | 48.2 | 55.6 | 52.6 | 52.18 | |
| InternVL3.5Size=8B, Modality=V2026.02 | 47.9 | 64.8 | 59.5 | 57.4 | |
| Qwen2.5-OmniSize=7B2026.01 | 46.9 | 66.1 | 60 | 57.7 | |
| Qwen2.5-OmniSize=7B, Modality=A + V2026.02 | 46.9 | 66.1 | 60 | 57.7 | |
| Qwen2.5-OmniParameters=7B2026.06 | 46.9 | 66.1 | 60 | 57.7 | |
| Qwen2.5-VLSize=7B, Modality=V2026.02 | 46.6 | 69.1 | 62.3 | 59.3 | |
| HumanOmniV2Size=7B, Modality=A + V2026.02 | 45.6 | 66.3 | 59.5 | 57.1 | |
| HumanOmniV2Parameters=7B2026.06 | 45.6 | 66.3 | 59.5 | 57.1 | |
| MiniCPM-o-2.6Size=8B, Modality=A + V2026.02 | 38.6 | 68.5 | 57.7 | 54.9 |