Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Video Question Answering on Video-MME w.o. sub. Medium
Loading...
70.8
Accuracy
Qwen2-VL + ReQuest
34.92
44.235
53.55
62.865
Jul 2, 2026
Accuracy
Updated 22d ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen2-VL + ReQuest
LLM Size=8B, #Frames=≤...
2026.07
70.8
Qwen2-VL
LLM Size=8B, #Frames=5...
2026.07
70.1
LLaVA-Video + ReQuest
LLM Size=7B, #Frames=32
2026.07
64.1
NVILA
LLM Size=8B, #Frames=1024
2026.07
62.2
VideoTree
LLM Size=–, #Frames=–
2026.07
59.9
LLaVA-Video
LLM Size=7B, #Frames=3...
2026.07
59.3
LLaVa-OneVision + ReQuest
LLM Size=7B, #Frames=3...
2026.07
58.8
LongVU
LLM Size=7B, #Frames=1fps
2026.07
58.2
LLaVa-OneVision
LLM Size=7B, #Frames=3...
2026.07
56.6
Frame-Voyager
LLM Size=7B, #Frames=8
2026.07
56.3
LLoVi
LLM Size=–, #Frames=–
2026.07
53.2
VideoLLaMA2
LLM Size=7B, #Frames=16
2026.07
45.4
Chat-UniVi-V1.5
LLM Size=7B, #Frames=64
2026.07
40.3
Video-LLaVA
LLM Size=7B, #Frames=8
2026.07
38
VideoChat2
LLM Size=7B, #Frames=16
2026.07
37
ShareGPT4Video
LLM Size=8B, #Frames=16
2026.07
36.3
Feedback
Search any
task
Search any
task