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

Multimodal Hallucination Evaluation on HallBench

65.2Score

GPT-5-Thinking

Updated 19d ago

Evaluation Results

Method	Links
GPT-5-Thinking 2025.09		65.2
Gemini-2.5-Pro 2025.09		64.1
VAPO-Thinker-7B 2025.09		57.4
VLAA-Thinker-7B 2025.09		54.7
R1-OneVision-7B 2025.09		52.5
Qwen2.5-VL-7B 2025.09		50
Vision-R1-7B 2025.09		49.5
VAPO-Thinker-3B 2025.09		49.5
InternVL2.5-8B 2025.09		49