Our new X account is live! Follow @wizwand_team for updates

Video Understanding on Perception (test)

82.7Accuracy

PLM-8B

Updated 3d ago

Evaluation Results

Method	Links
PLM-8B 2026.01		82.7	-
Molmo2-8B 2026.01		82.1	-
Molmo2-4B 2026.01		81.3	-
Eagle2.5-8B 2026.01		81	-
Molmo2-O-7B 2026.01		79.6	-
GPT-5 2026.01		79.4	-
PLM-3B 2026.01		79.3	-
Gemini 2.5 Pro 2026.01		78.4	-
Gemini 3 Pro 2026.01		77.6	-
VideoChat-Flash-7B 2026.01		76.5	-
Gemini 2.5 Flash 2026.01		74.7	-
GLM-4.1V-9B 2026.01		74.2	-
InternVL3.5-8B 2026.01		72.7	-
Qwen3-VL-8B 2026.01		72.7	-
GPT-5 mini 2026.01		72	-
MiniCPM-V-4.5-8B 2026.01		70.9	-
Qwen3-VL-4B 2026.01		70.7	-
LLaVA-Video-7B 2026.01		68.8	-
InternVL3.5-4B 2026.01		68.1	-
Qwen2-VL-72B 2024.09		68	-
Previous SoTA 2024.09		66.9	-
LLaVA-OV-72B 2024.12		66.9	-
Claude Sonnet 4.5 2026.01		64.3	-
Keye-VL-1.5-8B 2026.01		64.2	-
Qwen2-VL-7B 2024.09		62.3	-
LLaVA-OV-72B 2024.12		62.3	-
Qwen2-VL-7B-Ins. 2024.12		62.3	-
Llama 3-V 70B 2024.07		60.8	-
MAmmoTH-VL-8B 2024.12		59.3	-
InternVL-2-8B 2024.12		57.4	-
LLaVA-OV-7B 2024.12		57.1	-
LLaVA-OV-7B 2024.12		54.9	-
Gemini 1.0 Ultra 2024.07		54.7	-
Qwen2-VL-2B 2024.09		53.9	-
Llama 3-V 8B 2024.07		53.8	-
Gemini 1.0 Pro 2024.07		51.1	-
Video-LaVIT 2024.02		47.9	-
VideoChat2 2024.02		47.3	-
BLIP-2 2024.02		39.2	-
Flamingo 2024.02		33.5	-
Qwen2.5VL 3B 2025.12		-	67
CASA⊕ Q2.5VL 2025.12		-	58.1