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

Long-form Video Understanding on LongVideoBench (Accuracy)

82.3Accuracy

VideoChat-M1

Updated 16d ago

Evaluation Results

Method	Links
VideoChat-M1 2025.11		82.3
GPT-4o 2026.05		81.3
Gemini 2.5 Pro 2025.11		78.7
Seed1.5-VL 2025.05		74.4
Seed1.5-VL 2025.05		74
Seed 1.5VL 2025.11		74
DeepVideoDiscovery 2025.11		71.6
NVILA 2025.11		70.1
InternVL3.5-241B-A28B 2026.01		67.1
InternVL-3.5-241B 2025.11		67.1
GPT-4o 2024.09		66.7
GPT-4o 2025.05		66.7
GPT-4o 2026.02		66.7
GPT-4o 2025.12		66.7
GPT-4o 2025.12		66.7
GPT-4o 2025.11		66.7
GPT4o 2025.11		66.7
VideoSeeker-8B 2026.05		66.5
Eagle-2.5-8B 2025.11		66.4
InternVL-3-78B 2025.11		65.7
InternVL-3.5-78B 2025.11		65.7
VideoChat-A1 2025.11		65.4
VideoRAG-72B 2025.11		65.4
LLAVA-Video-72B 2025.11		64.9
VideoChat-Flash-7B 2025.11		64.7
Qwen3-VL-8B 2026.05		64.6
Gemini-1.5-Pro 2024.09		64.4
Aria-28B 2025.11		64.2
VideoSeeker-4B 2026.05		64.2
Gemini-1.5-Pro 2026.02		64
Gemini-1.5-pro 2025.12		64
Gemini-1.5-pro 2025.12		64
Gemini 1.5 Pro 2025.11		64
Gemini-1.5-Pro 2025.11		64
Qwen3-VL-8B-Thinking + MVP 2026.01		63.8
InternVL-2.5-78B 2025.11		63.6
Qwen3-VL-8B-Thinking 2026.01		62.9
VideoChat-R1.5-7B 2025.11		62.6
Qwen3-VL-4B 2026.05		62.6
Oryx 2024.09		62.2
InternVL3.5 2026.02		62.1
InternVL-3.5-8B 2025.11		62.1
Oryx-1.5 2024.09		62
Qwen3-VL-8B-Thinking + MVP 2026.01		62
LLaVA-Video 2026.02		61.9
Demo-ICL (SFT) 2026.02		61.6
VidLaDA 2026.01		61.4
LLaVA-OneVision 2024.09		61.3
LLaVA-OneVision 2026.02		61.3
LLaVA-ov-72B 2025.11		61.3
Demo-ICL 2026.02		61.2
InternVL3.5-8B + MVP 2026.01		61
VideoXL2-8B 2025.11		61
GPT-4V 2024.09		60.7
Qwen2.5-VL 2026.02		60.7
Ola-Video (Base) 2026.02		60.7
Qwen2.5-VL-72B 2026.01		60.7
Qwen2.5-VL-72B 2025.11		60.7
InternVideo2.5-7B 2025.11		60.6
Hour-LLaVA 2025.11		60.4
Qwen2.5-VL 2026.01		60.2
InternVL2.5 2026.01		60
InternVL-2.5-8B 2025.11		60
VideoLLaMA3 2026.02		59.8
Qwen3-VL-8B-Thinking 2026.01		59.6
InternVL3.5-8B + MVP 2026.01		59.5
Streamo-7B 2025.12		59.2
Streamo-7B 2025.12		59.2
StreamingVLM-7B 2025.12		59
StreamingVLM-7B 2025.12		59
InternVL-3-8B 2025.11		58.8
LLaDA-V 2026.01		58.6
Qwen2.5-VL-7B-Instruct + MVP 2026.01		58.6
GPT-4o 2026.02		58.5
LLaVA-Video 2026.02		58.2
LLaVA-Video 2026.01		58.2
LLaVA-Video-7B 2025.11		58.2
LongRL-7B 2025.11		58.1
LongVILA-R1 2025.11		58
REVISOR (Ours) 2025.11		57.5
Qwen2.5-VL* 2025.11		57.4
LongVILA-7B 2025.11		57.1
LLaVA-OneVision 2026.01		56.5
Streamo-2B 2025.12		56.5
Qwen2.5-VL⋆ 2025.11		56.5
LLaVA-OneVision 2025.11		56.4
VL-Rethinker 2025.11		56.4
Oryx-1.5 2024.09		56.3
LLaVA-OneVision 2026.02		56.3
Jigsaw-7B 2026.01		56.2
Streamo-3B 2025.12		56.2
Streamo-3B 2025.12		56.2
Streamo-4B 2025.12		56.1
Qwen2.5-VL 2026.02		56
Qwen2.5-VL-7B 2025.12		56
Qwen2.5-VL-7B 2025.12		56
Qwen2.5-VL-7B 2025.11		56
Qwen2.5-VL-7B-Instruct + MVP 2026.01		55.9
Vamba 2025.11		55.9
InternVL3.5-8B 2026.01		55.6

Showing 100 of 135 rows