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

Visual Reasoning on HR-Bench 4K

0.819Overall Score

P2R-4B

Updated 23d ago

Evaluation Results

Method
P2R-4B 2026.07	0.819	0.923	0.715
P2R-8B 2026.07	0.815	0.935	0.715
Qwen3-VL-Instruct-32B 2026.07	0.786	0.86	0.713
SubagentVL 2025.12	0.77	0.933	0.608
Thyme-7B 2026.07	0.77	0.91	0.63
DeepEyes 2025.12	0.751	0.92	0.583
DeepEyes 2025.12	0.751	0.913	0.59
DeepEyes-7B 2026.07	0.751	0.913	0.59
P2R-2B 2026.07	0.751	0.898	0.601
Qwen3-VL-Instruct-8B 2026.07	0.748	0.81	0.685
Qwen2.5-VL 2025.12	0.739	0.898	0.58
Qwen3-VL-Instruct-4B 2026.07	0.738	0.818	0.66
UniVLR 2026.05	0.733	0.86	0.605
RIS 2026.05	0.7323	0.8633	0.6012
ERGO 2025.09	0.73	-	-
PixelReasoner 2026.05	0.729	0.86	0.603
PixelReasoner-7B 2026.07	0.729	0.86	0.603
CoVT 2026.05	0.719	0.855	0.583
Monet 2026.05	0.719	0.893	0.545
CoVT 2026.05	0.719	0.842	0.595
RIS+VLPO 2026.05	0.7179	0.8283	0.6075
DeepEyes 2026.05	0.713	0.838	0.588
Qwen2.5-VL-7B-Inst. 2025.09	0.711	-	-
LVR 2026.05	0.7088	0.8325	0.575
Qwen2.5-VL-7B+GLSD 2026.05	0.7058	0.833	0.5786
Qwen3-VL-Instruct-2B 2026.07	0.704	0.81	0.598
SkiLa 2026.05	0.703	0.847	0.557
MiniO3 2025.09	0.7	-	-
Monet 2026.05	0.6997	0.8401	0.5593
LVR 2026.05	0.699	0.847	0.557
MGPO 2025.09	0.698	-	-
ZoomEye 2025.12	0.696	0.843	0.55
ZoomEye-7B 2026.07	0.696	0.843	0.55
Qwen2.5-VL-7B + vanilla SFT 2026.05	0.691	0.813	0.57
Qwen2.5-VL-7B 2026.05	0.69	0.858	0.522
Qwen2.5-VL 2025.12	0.688	0.852	0.522
Qwen2.5-VL-7B 2026.05	0.683	0.806	0.5603
ERGO 2025.09	0.671	-	-
PixelReasoner 2025.09	0.669	-	-
VisionThink 2025.09	0.669	-	-
PixelReasoner 2025.09	0.665	-	-
TreeVGR 2025.09	0.664	-	-
VisionThink 2025.09	0.661	-	-
DeepEyes 2025.09	0.66	-	-
Qwen2.5-VL-7B-Inst. 2025.09	0.656	-	-
DeepEyes 2025.09	0.644	-	-
MiniO3 2025.09	0.628	-	-
MGPO 2025.09	0.628	-	-
TreeVGR 2025.09	0.624	-	-
GPT-4o 2025.12	0.59	0.7	0.48
GPT-4o 2026.05	0.59	0.7	0.48
GPT-4o 2026.07	0.59	0.7	0.48
Qwen2.5-VL-7B-Inst. 2025.09	0.57	-	-
GPT-4o 2026.05	0.547	0.6493	0.4451