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

Visual Reasoning on HR-Bench 8K

72.6Overall Score

DeepEyes

Updated 21d ago

Evaluation Results

Method	Links
DeepEyes 2025.12		72.6	86.8	58.5
SubagentVL 2025.12		72.6	87	58.3
DeepEyes 2025.12		71	85	57
Qwen2.5-VL 2025.12		70.4	84.5	56.3
ERGO 2025.09		69.9	-	-
CoVT 2026.05		69.7	85.4	54
ZoomEye 2025.12		69.3	85.5	50
UniVLR 2026.05		68.8	78.8	58.8
RIS 2026.05		68.52	79.05	57.98
CoVT 2026.05		68.4	79.3	57.5
Qwen2.5-VL-7B-Inst. 2025.09		67.1	-	-
Monet 2026.05		67.01	78.59	55.43
PixelReasoner 2026.05		66.9	80	54.3
LVR 2026.05		66.9	77.7	56.2
ERGO 2025.09		66.1	-	-
Qwen2.5-VL-7B 2026.05		66	80.3	51.8
MiniO3 2025.09		65.9	-	-
VisionThink 2025.09		65.8	-	-
Qwen2.5-VL 2025.12		65.3	78.8	51.8
DeepEyes 2026.05		65.1	77	53.3
Qwen2.5-VL-7B+GLSD 2026.05		64.7	74.69	54.7
Qwen2.5-VL-7B 2026.05		64.33	74.42	54.24
Qwen2.5-VL-7B + vanilla SFT 2026.05		63.6	73.3	54
LVR 2026.05		63.5	75	52
Monet 2026.05		63.5	76.5	50.5
RIS+VLPO 2026.05		63.05	72.67	53.42
SkiLa 2026.05		62.9	77.5	48.3
PixelReasoner 2025.09		61.3	-	-
MGPO 2025.09		61.1	-	-
TreeVGR 2025.09		60.4	-	-
VisionThink 2025.09		60.1	-	-
DeepEyes 2025.09		60	-	-
PixelReasoner 2025.09		59.9	-	-
DeepEyes 2025.09		58.3	-	-
MiniO3 2025.09		57.3	-	-
MGPO 2025.09		57.3	-	-
Qwen2.5-VL-7B-Inst. 2025.09		56.5	-	-
GPT-4o 2025.12		55.5	62	49
GPT-4o 2026.05		55.5	62	49
TreeVGR 2025.09		54.4	-	-
GPT-4o 2026.05		51.12	57.08	45.12
Qwen2.5-VL-7B-Inst. 2025.09		49.9	-	-