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

Scientific Multimodal Reasoning on ScienceOlympiad

41.38Accuracy

GPT-5

Updated 1mo ago

Evaluation Results

Method	Links
GPT-5 2026.04		41.38
Qwen3-VL-235B-A22B-Thinking 2026.04		36.47
Intern-S1 2026.04		36.47
S1-VL-32B-RL 2026.04		33.5
S1-VL-32B-SFT 2026.04		33
Gemini 2.5 Flash 2026.04		28.57
Gemini 2.5 Pro 2026.04		24.13
Qwen3-VL-32B-Thinking 2026.04		22.35
Intern-S1-mini 2026.04		11.82
Thyme-VL 2026.04		1.97