Share your thoughts, 1 month free Claude Pro on usSee more

Long-video Question Answering on Long VideoBench (val)

73.2Accuracy

GOPAgen

Updated 1mo ago

Evaluation Results

Method	Links
GOPAgen 2026.06		73.2
GPT-5 2025.12		72.6
Deep Video Discovery 2026.06		71.6
VideoLucy 2026.06		70.2
OpenAI o3 2026.06		67.5
AdaReTaKe 2026.06		67
GPT-4o 2025.01		66.7
GPT-4o 2025.02		66.7
GPT-4o 2025.12		66.7
GPT-4o 2026.06		66.7
Gemini-1.5-Pro 2025.01		64
Gemini-1.5-Pro 2025.02		64
Gemini-1.5-Pro 2025.12		64
Gemini-1.5-Pro 2026.06		64
InternVL2.5-78B 2025.01		63.6
InternVL2.5-78B 2026.06		63.6
LLaVA-Video + OneClip-RAG 2025.12		62.9
LLaVA-Video w/ AKS 2025.02		62.7
AKS 2025.12		62.7
InternVL3.5 2025.12		62.1
LLaVA-Video-72B 2025.01		61.9
Gemini-1.5-Flash 2025.02		61.6
GPT-4V 2025.02		61.3
Qwen2.5-VL-72B 2026.06		60.7
Qwen2-VL w/ AKS 2025.02		60.5
InternVL2.5-8B 2025.01		60
VideoNSA 2025.12		60
mPLUG-Owl3 2026.06		59.8
LLaVA-OV w/ AKS 2025.02		59.3
LLaVA-Video 2025.12		59.3
LLaVA-Video 2025.02		58.9
Tarsier2-7B 2025.01		58.6
LLaVA-Video-7B 2025.01		58.2
NVILA 2025.12		57.7
Qwen2.5-VL 2025.12		56
Qwen2-VL-7B 2025.01		55.6
Qwen2-VL 2025.02		55.5
MiniCPM-V 2.6 2025.02		54.9
LLaVA-OV 2025.02		54.8
Tarsier-34B 2025.01		54.2
PLLAVA 2025.02		53.2
mPLUG-Owl3 2025.12		52.1
Video-XL 2025.12		50.7
LongVA 2025.12		47.8
Tarsier-7B 2025.01		39.8
VideoLLaVA 2025.02		39.1