List your company

Multimodal Reasoning on SPD-Faith Bench Easy 1.0

5Contradiction RateGPT-4o
3.6212.93522.2531.565Feb 8, 2026

Evaluation results

MethodLinks
GPT-4oModel Type=Proprietary2026.025
GLM-4.5VModel Type=Open-Source2026.028
Qwen3-VL-235B-A22BModel Type=Open-Source, Parameters=235B-A22B2026.0210.5
Qwen3-VL-32BModel Type=Open-Source, Parameters=32B2026.0216
Gemini-2.5-ProModel Type=Proprietary2026.0216.5
Claude-4.5-HaikuModel Type=Proprietary2026.0216.5
Qwen2.5-VL-72BModel Type=Open-Source, Parameters=72B2026.0219
InternVL2.5-38BModel Type=Open-Source, Parameters=38B2026.0219.2
Qwen2.5-VL-7BModel Type=Open-Source, Parameters=7B2026.0223
MiniCPM-V-2.6Model Type=Open-Source2026.0223
InternVL2.5-8BModel Type=Open-Source, Parameters=8B2026.0232.5
DeepSeek-VL2Model Type=Open-Source2026.0239.5