Long-video understanding on Video-MME
88.6Overall ScoreGemini 3 Pro
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 3 ProModel Type=Proprietary2026.06 | 88.6 | — | — | — | — | — | — | — | — | — | |
| Gemini 2.5 ProModel Type=Proprietary2026.06 | 87.8 | — | — | — | — | — | — | — | — | — | |
| Gemini-2.5-ProLLM Size=-2026.03 | 84.3 | — | — | — | — | — | — | — | — | — | |
| Gemini 2.5 FlashModel Type=Proprietary2026.06 | 84.2 | — | — | — | — | — | — | — | — | — | |
| GPT-5Model Type=Proprietary2026.06 | 83.3 | — | — | — | — | — | — | — | — | — | |
| GPT-5 miniModel Type=Proprietary2026.06 | 77.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ AdaptTokenLLM Size=72B2026.03 | 76.1 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-ProEvaluation Protocol=Zero-shot2025.07 | 75 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-ProFrames=-, LLM Size=-2026.06 | 75 | — | — | — | — | — | — | — | — | 67.4 | |
| Qwen2.5-VL w/ FlexSelectLLM Size=72B2026.03 | 74.4 | — | — | — | — | — | — | — | — | — | |
| Claude Sonnet 4.5Model Type=Proprietary2026.06 | 74.2 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL w/ AdaptTokenLLM Size=8B2026.03 | 73.8 | — | — | — | — | — | — | — | — | — | |
| InternVideo3Model Type=Open-Weight2026.06 | 73.8 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ AdaReTAKELLM Size=72B2026.03 | 73.5 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLLLM Size=72B2026.03 | 73.4 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-ProLLM Size=-2026.03 | 73.2 | — | — | — | — | — | — | — | — | — | |
| Keye-VL-1.5-8BModel Type=Open-Weight, Parameters=8B2026.06 | 73 | — | — | — | — | — | — | — | — | — | |
| Eagle2.5-8BModel Type=Open-Weight, Parameters=8B2026.06 | 72.4 | — | — | — | — | — | — | — | — | — | |
| GPT-4oLLM Size=-2026.03 | 71.9 | — | — | — | — | — | — | — | — | — | |
| GPT-4oEvaluation Protocol=Zero-shot2025.07 | 71.9 | — | — | — | — | — | — | — | — | — | |
| GPT-4oFrames=-, LLM Size=-2026.06 | 71.9 | — | — | — | — | — | — | — | — | 65.3 | |
| Qwen3-VLLLM Size=8B2026.03 | 71.4 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-8BModel Type=Open-Weight, Parameters=8B2026.06 | 71.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ AdaptTokenLLM Size=7B2026.03 | 70.5 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-FlashEvaluation Protocol=Zero-shot2025.07 | 70.3 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-FlashFrames=-, LLM Size=-2026.06 | 70.3 | — | — | — | — | — | — | — | — | 61.1 | |
| Molmo2-8BModel Type=Open-Weight, Parameters=8B2026.06 | 69.9 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ AdaptToken-LiteLLM Size=7B2026.03 | 69.8 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ AdaptTokenLLM Size=8B2026.03 | 68.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ FlexSelectLLM Size=7B2026.03 | 68.2 | — | — | — | — | — | — | — | — | — | |
| GLM-4.1V-9BModel Type=Open-Weight, Parameters=9B2026.06 | 68.2 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ AdaptToken-LiteLLM Size=8B2026.03 | 68.1 | — | — | — | — | — | — | — | — | — | |
| MiniCPM-V-4.5-8BModel Type=Open-Weight, Parameters=8B2026.06 | 67.9 | — | — | — | — | — | — | — | — | — | |
| MMProLong2026.05 | 67.78 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ AdaReTAKELLM Size=7B2026.03 | 67.7 | — | — | — | — | — | — | — | — | — | |
| ViLAMPLLM Size=7B2026.03 | 67.5 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ FlexSelectLLM Size=8B2026.03 | 67 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ TimeSearch-RLLM Size=7B2026.03 | 66.6 | — | — | — | — | — | — | — | — | — | |
| VideoLLaMA3LLM Size=7B2026.03 | 66.2 | — | — | — | — | — | — | — | — | — | |
| VideoLLaMA3LLM=Qwen2.5-7B, Frames=1fps, Evaluation Protocol=Zero-shot2025.07 | 66.2 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-8BModel Type=Open-Weight, Parameters=8B2026.06 | 66 | — | — | — | — | — | — | — | — | — | |
| TPOLLM Size=7B2026.03 | 65.6 | — | — | — | — | — | — | — | — | — | |
| MiMo-VL w/ Q-FoldFrames=32, LLM Size=7B2026.06 | 65.6 | — | — | — | — | — | — | — | — | 54.3 | |
| Qwen2.5-VL w/ SeViCESLLM Size=7B2026.03 | 65.5 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ TriumphLLM Size=8B2026.03 | 65.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLLLM Size=7B2026.03 | 65.4 | — | — | — | — | — | — | — | — | — | |
| VideoChat-Flash-7BModel Type=Open-Weight, Parameters=7B2026.06 | 65.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLLLM=Qwen2.5-7B, Evaluation Protocol=Zero-shot2025.07 | 65.1 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B2026.05 | 65.1 | — | — | — | — | — | — | — | — | — | |
| InternVideo2.5-7BModel Type=Open-Weight, Parameters=7B2026.06 | 65.1 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ SeViCESLLM Size=8B2026.03 | 64.7 | — | — | — | — | — | — | — | — | — | |
| ByteVideoLLMLLM Size=14B2026.03 | 64.6 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5 w/ ZoomVLLM Size=8B2026.03 | 64.4 | — | — | — | — | — | — | — | — | — | |
| NVILALLM Size=8B2026.03 | 64.2 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5LLM Size=8B2026.03 | 64.2 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ ZoomVLLM Size=7B2026.03 | 63.6 | — | — | — | — | — | — | — | — | — | |
| LLaVA-Video-7BModel Type=Open-Weight, Parameters=7B2026.06 | 63.3 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-REGATELLM=Qwen3-14B, Frames=16, Tokens=2.32B (↓ 41.41%), Evaluation Protocol=Zero-shot2025.07 | 63 | — | — | — | — | — | — | — | — | — | |
| MiMo-VLFrames=32, LLM Size=7B2026.06 | 62.5 | — | — | — | — | — | — | — | — | 52.6 | |
| InternVL3.5LLM=Qwen3-14B, Frames=16, Tokens=3.96B, Evaluation Protocol=Zero-shot2025.07 | 62.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL w/ Q-FoldFrames=32, LLM Size=7B2026.06 | 62.1 | — | — | — | — | — | — | — | — | 52.6 | |
| Qwen2-VL w/ Q-FoldFrames=32, LLM Size=7B2026.06 | 61.7 | — | — | — | — | — | — | — | — | 51.9 | |
| Qwen2.5-VLFrames=32, LLM Size=7B2026.06 | 61.5 | — | — | — | — | — | — | — | — | 50.8 | |
| LLaVA-OneVision-1.5 w/ Q-FoldFrames=32, LLM Size=8B2026.06 | 61.5 | — | — | — | — | — | — | — | — | 52.2 | |
| LongVUFrames=1 FPS, LLM Size=7B2026.06 | 60.6 | — | — | — | — | — | — | — | — | — | |
| LLaVA-OneVision-1.5Frames=32, LLM Size=8B2026.06 | 60.5 | — | — | — | — | — | — | — | — | 51.1 | |
| LLaVA-NeXT-VideoLLM=Qwen2-32B, Frames=32, Evaluation Protocol=Zero-shot2025.07 | 60.2 | — | — | — | — | — | — | — | — | — | |
| VILA1.5LLM=Llama-2-40B, Frames=8, Evaluation Protocol=Zero-shot2025.07 | 60.1 | — | — | — | — | — | — | — | — | — | |
| LongVILAFrames=128, LLM Size=7B2026.06 | 60.1 | — | — | — | — | — | — | — | — | — | |
| GPT-4VFrames=-, LLM Size=-2026.06 | 59.9 | — | — | — | — | — | — | — | — | 53.5 | |
| Qwen2-VL w/ AKSFrames=32, LLM Size=7B2026.06 | 59.9 | — | — | — | — | — | — | — | — | — | |
| Qwen2-VL w/ FOCUSFrames=32, LLM Size=7B2026.06 | 59.7 | — | — | — | — | — | — | — | — | — | |
| Qwen2-VL w/ TSPO†Frames=32, LLM Size=7B2026.06 | 59.6 | — | — | — | — | — | — | — | — | — | |
| mPLUG-Owl3LLM Size=7B2026.03 | 59.3 | — | — | — | — | — | — | — | — | — | |
| mPLUG-Owl3Frames=128, LLM Size=7B2026.06 | 59.3 | — | — | — | — | — | — | — | — | 50.1 | |
| Oryx-1.5Frames=64, LLM Size=7B2026.06 | 58.8 | — | — | — | — | — | — | — | — | 51.2 | |
| Qwen2-VL w/ Q-FrameFrames=4 + 8 + 32, LLM Size=7B2026.06 | 58.3 | — | — | — | — | — | — | — | — | 48.3 | |
| LLaVA-OneVisionLLM=Qwen2-7B, Frames=32, Evaluation Protocol=Zero-shot2025.07 | 58.2 | — | — | — | — | — | — | — | — | — | |
| VAMBAFrames=1024, LLM Size=13B2026.06 | 57.8 | — | — | — | — | — | — | — | — | — | |
| TimeMarkerFrames=128, LLM Size=8B2026.06 | 57.3 | — | — | — | — | — | — | — | — | 46.4 | |
| Qwen2-VLFrames=32, LLM Size=7B2026.06 | 57.3 | — | — | — | — | — | — | — | — | 47.3 | |
| Video-XLFrames=128, LLM Size=7B2026.06 | 55.5 | — | — | — | — | — | — | — | — | 49.2 | |
| VideoLLaMA2-REGATELLM=Qwen2-7B, Frames=16, Tokens=49.27M (↓ 41.22%), Evaluation Protocol=Zero-shot2025.07 | 54.5 | — | — | — | — | — | — | — | — | — | |
| VideoLLaMA2LLM=Qwen2-7B, Frames=16, Tokens=83.82M, Evaluation Protocol=Zero-shot2025.07 | 53.7 | — | — | — | — | — | — | — | — | — | |
| LongVAFrames=128, LLM Size=7B2026.06 | 52.6 | — | — | — | — | — | — | — | — | 46.2 | |
| Video-CCAMFrames=96, LLM Size=9B2026.06 | 50.3 | — | — | — | — | — | — | — | — | 39.6 | |
| Video-LLaVALLM=Vicuna-7B, Frames=8, Evaluation Protocol=Zero-shot2025.07 | 39.9 | — | — | — | — | — | — | — | — | — | |
| VideoChat2-REGATELLM=Mistral-7B, Frames=16, Tokens=2.22B (↓ 43.51%), Evaluation Protocol=Zero-shot2025.07 | 32.7 | — | — | — | — | — | — | — | — | — | |
| VideoChat2LLM=Mistral-7B, Frames=16, Tokens=3.93B, Evaluation Protocol=Zero-shot2025.07 | 26 | — | — | — | — | — | — | — | — | — | |
| LLaMA-VIDLLM=Llama-2-7B, Frames=1fps, Evaluation Protocol=Zero-shot2025.07 | 25.9 | — | — | — | — | — | — | — | — | — | |
| AdaReTaKe2026.05 | — | — | — | — | — | — | — | 63.1 | — | — | |
| Gemini 1.5 FlashCategory=Proprietary Models2026.03 | — | 78.8 | 79.8 | 68.8 | 74.7 | 61.1 | 68.8 | 70.3 | 75 | — | |
| Gemini 1.5 ProCategory=Proprietary Models2026.03 | — | 81.7 | 84.5 | 74.3 | 81 | 67.4 | 77.4 | 75 | 81.3 | — | |
| GPT-4oCategory=Proprietary Models2026.03 | — | 80 | 82.8 | 70.3 | 76.6 | 65.3 | 72.1 | 71.9 | 77.2 | — | |
| GPT-4VCategory=Proprietary Models2026.03 | — | 70.5 | 73.2 | 55.8 | 59.7 | 53.5 | 56.9 | 59.9 | 63.3 | — | |
| LLaVA-Video-72BCategory=Open-Sourced MLLMs2026.03 | — | 81.4 | 82.8 | 68.9 | 75.6 | 61.5 | 72.5 | 70.6 | 76.9 | — | |
| Long-LLaVA-7BCategory=Open-Sourced MLLMs2026.03 | — | 61.9 | 66.2 | 51.4 | 54.7 | 45.4 | 50.3 | 52.9 | 57.1 | — | |
| LongVA-7BCategory=Open-Sourced MLLMs2026.03 | — | 61.1 | 61.6 | 50.4 | 53.6 | 46.2 | 47.6 | 52.6 | 54.3 | — | |
| MM-MemBase Model=Qwen3-VL-8B, Category=Agent-based Systems2026.03 | — | 81.5 | 82.8 | 69.6 | 75.8 | 66.1 | 75.7 | 72.4 | 78.1 | — | |
| Qwen2-VL-72BCategory=Open-Sourced MLLMs2026.03 | — | 80.1 | 82.2 | 71.3 | 76.8 | 62.2 | 74.3 | 71.2 | 77.8 | — |