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

Visual Reasoning on BLINK

85.2Accuracy

Qwen3-VL-8B-Inst.

Updated 18d ago

Evaluation Results

Method	Links
Qwen3-VL-8B-Inst. 2026.03		85.2
Vlaser-8B 2026.03		84.9
RoboBrain2.5-8B 2026.03		84.3
InternVL3.5-8B 2026.03		84.1
ACE-Brain-0-8B 2026.03		83.9
InternVL3-8B 2026.03		82.9
Qwen2.5-VL-7B-Inst. 2026.03		82.5
Gemini-2.5-Pro 2026.03		81.8
RoboBrain2.0-7B 2026.03		81.4
GPT-5-mini 2025.12		81
VeBrain-7B 2026.03		79.7
Claude-4-Sonnet 2026.03		78.1
GPT-4o 2026.03		77.9
GPT-5-mini 2025.12		76
Qwen3-VL-2B-Inst. 2026.03		74.9
Gemini-2.5-Pro 2026.01		70.6
GRIT 2025.12		70.3
VALOR 2025.12		69.2
VALOR 2025.12		69.2
VALOR 2025.12		69.2
ViGoRL 2025.12		68.4
GPT-4o 2026.01		68
GPT-4o 2025.09		68
VALOR-RL 2025.12		67.3
VALOR-RL 2025.12		67.3
Qwen3-VL 2026.05		66.6
o4-mini 2025.12		66.5
GPT-4o 2026.01		65.9
Claude-3.5-Haiku 2025.12		64.6
GPT-4o 2025.12		64.2
Qwen3-8B 2025.12		63.9
GPT-4o 2026.05		63.55
Gemini2.0-Flash 2025.09		63.5
Qwen3.5-VL 2026.05		63.5
SmoothOp-7B 2026.01		62.6
VST-7B-SFT 2026.01		62.1
Qwen3.5-VL 2026.05		62
Gemini-2.5-Flash 2025.12		61.5
RIS+VLPO 2026.05		60.95
RIS 2026.05		60.6
VST-3B-SFT 2026.01		59.1
InternVL3.5 2026.05		57.8
CoVT 2026.07		57.49
ProLaViT (Base) 2026.07		57.49
Gemma-3-12B 2025.12		57.4
CoVT 2026.05		57.4
ProLaViT (Full) 2026.07		57.23
VisionZero-Qwen-7B (Real-World) 2025.09		57.2
Laser 2026.01		56.92
SpatialLadder-3B 2026.01		56.9
SmoothOp-3B 2026.01		56.9
Pelican-VL-7B 2026.03		56.8
VisionZero-Qwen-7B (Chart) 2025.09		56.8
LVR 2026.05		56.79
Monet 2026.05		56.7
CoT SFT 2026.07		56.54
Qwen2.5-VL-7B 2026.01		56.4
ProLaViT + L_div^margin 2026.07		56.33
One-step Latent Pred. 2026.07		56.28
Qwen2.5-VL-7B+GLSD 2026.05		56.25
ViLaSR-7B 2026.01		56.2
VisionZero-Qwen-7B (CLEVR) 2025.09		56
Qwen3-VL 2026.05		55.7
ViGaL-Snake+Rotation 2025.09		55.6
SFT 2026.07		55.6
VL-Rethinker 2026.01		55.55
InternVL3-8B 2026.01		55.5
SpaceR-7B 2026.01		55.4
Qwen2.5-VL-7B 2025.09		55.2
Qwen2.5-VL-7B 2026.05		54.94
InternVL3.5-8B 2026.01		54.81
Qwen2.5-VL-Instruct 2026.07		54.49
Llama-3.2-11B 2025.12		54.3
MM-Eureka-Qwen-7B 2025.09		54
Qwen2.5-VL-7B 2026.01		53.6
LVR 2026.01		53.6
LVR 2026.07		53.02
VLAA-Thinker-7B 2025.09		53
Vision-R1 2026.01		52.71
LLaVA 2025.12		52.7
PAPO 2026.01		52.66
InternVL3.5 2026.05		52.6
Gemini-2.0-Flash 2025.12		51.7
VISTA-SFT 2026.05		51.23
DeepEyes 2026.01		51.08
Monet 2026.01		50.71
OpenVLThinker-7B 2025.09		49.9
STaR 2026.05		49.78
Ours w/o Masking 2025.12		49.6
SFT-Seed 2026.05		49.35
LLaVA-OneVision 2026.01		49.34
VIRAL 2025.12		49.3
JARVIS 2025.12		49.3
R1-OneVision-7B 2025.09		48.7
JARVIS 2025.12		48.3
LLava-OneVision-7B 2026.01		48.2
ReSTEM 2026.05		47.62
Qwen2.5-VL-3B 2026.01		47.6
LLaVA 2025.12		46.8
Datology 2026.05		45.9

Showing 100 of 116 rows