Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Long Video Question Answering on Video-MME w/o subtitles
Loading...
0.818
Accuracy
GPT-5
0.51432
0.59316
0.672
0.75084
Dec 9, 2025
Accuracy
Updated 3d ago
Evaluation Results
Method
Method
Links
Accuracy
GPT-5
LLM=N/A, # Frames=N/A
2025.12
0.818
Gemini-1.5-Pro
LLM=N/A, # Frames=N/A
2025.12
0.75
GPT-4o
LLM=N/A, # Frames=N/A
2025.12
0.719
InternVL3.5
LLM=8B, # Frames=64
2025.12
0.66
AKS
LLM=7B, # Frames=64
2025.12
0.653
Qwen2.5-VL
LLM=7B, # Frames=32
2025.12
0.651
LLaVA-Video + OneClip-RAG
LLM=7B, # Frames=64
2025.12
0.649
ByteVideoLLM
LLM=14B, # Frames=256
2025.12
0.646
NVILA
LLM=8B, # Frames=256
2025.12
0.642
LLaVA-Video
LLM=7B, # Frames=64
2025.12
0.633
LongVU
LLM=7B, # Frames=1fps
2025.12
0.606
mPLUG-Owl3
LLM=7B, # Frames=16
2025.12
0.593
Video-XL
LLM=7B, # Frames=128
2025.12
0.555
LongVA
LLM=7B, # Frames=256
2025.12
0.526
Feedback
Search any
task
Search any
task