InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning

Benchmark Model Rank Results
zero-shot-video-question-answer-on-egoschema-fullsetInternVideo3-8B#3Accuracy: 76.6
video-question-answering-on-tomatoInternVideo3-8B#5Accuracy: 37.4
video-question-answering-on-tempcompass-overallInternVideo3-8B#6Accuracy: 74.0
video-question-answering-on-mvbenchInternVideo3-8B#10Avg.: 75.0
video-question-answering-on-mlvuInternVideo3-8B#13Accuracy: 77.3
zero-shot-video-question-answer-on-video-mme-w-o-subsInternVideo3-8B#13Accuracy (%): 73.8
video-question-answering-on-lvbenchInternVideo3-8B#22Accuracy: 55.7