Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Multimodal Reasoning on SPD-Faith Bench Medium 1.0
Loading...
11.3
Contradiction Rate
Qwen2.5-VL-7B
10.232
17.441
24.65
31.859
Feb 8, 2026
Contradiction Rate
Updated 3d ago
Evaluation Results
Method
Method
Links
Contradiction Rate
Qwen2.5-VL-7B
Model Type=Open-Source...
2026.02
11.3
Qwen3-VL-32B
Model Type=Open-Source...
2026.02
12.3
MiniCPM-V-2.6
Model Type=Open-Source
2026.02
12.7
Qwen2.5-VL-72B
Model Type=Open-Source...
2026.02
13.2
GPT-4o
Model Type=Proprietary
2026.02
14.6
Qwen3-VL-235B-A22B
Model Type=Open-Source...
2026.02
15.2
GLM-4.5V
Model Type=Open-Source
2026.02
18.8
Gemini-2.5-Pro
Model Type=Proprietary
2026.02
21.9
InternVL2.5-38B
Model Type=Open-Source...
2026.02
22.4
InternVL2.5-8B
Model Type=Open-Source...
2026.02
29.7
Claude-4.5-Haiku
Model Type=Proprietary
2026.02
30.1
DeepSeek-VL2
Model Type=Open-Source
2026.02
38
Feedback
Search any
task
Search any
task