Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Visual Reasoning on VSP
Loading...
83.7
Accuracy
DeepLatent-RL-7B
9.3192
28.6296
47.94
67.2504
Jan 26, 2026
Feb 15, 2026
Mar 8, 2026
Mar 29, 2026
Apr 18, 2026
May 9, 2026
May 30, 2026
Accuracy
Updated 1d ago
Evaluation Results
Method
Method
Links
Accuracy
DeepLatent-RL-7B
Model=DeepLatent-RL-7B
2026.05
83.7
AdaReasoner 7B
Model Category=Tool-Pl...
2026.01
78.36
Mirage-7B
Model=Mirage-7B
2026.05
76
GPT 5 + Tools
Model Category=Tool-Pl...
2026.01
71.36
Claude 4 sonnet
Model Category=Closed-...
2026.01
56.27
GPT 5
Model Category=Closed-...
2026.01
55.64
Gemini 2.5 flash
Model Category=Closed-...
2026.01
53.55
Qwen 2.5 VL 72B + Tools
Model Category=Tool-Pl...
2026.01
45
Qwen 2.5 VL 72B
Model Category=Open-So...
2026.01
39.09
InternVL3 78B
Model Category=Open-So...
2026.01
35.09
Qwen 2.5 VL 32B
Model Category=Open-So...
2026.01
33.91
Qwen 2.5 VL 7B + Tools
Model Category=Tool-Pl...
2026.01
30.45
Qwen 2.5 VL 7B
Model Category=Open-So...
2026.01
28.09
Qwen 2.5 VL 3B
Model Category=Open-So...
2026.01
26.73
PixelReasoner
Model Category=Tool-Pl...
2026.01
24.55
Qwen2.5-VL-7B
Model=Qwen2.5-VL-7B
2026.05
13.5
DeepEyes
Model Category=Tool-Pl...
2026.01
12.18
Feedback
Search any
task
Search any
task