List your company

Multimodal Autoformalization on PHYX Mechanics

71.4Compile RateGPT-5
-2.85616.42235.754.978Jan 6, 2026

Evaluation results

MethodLinks
GPT-5Modality=Image2026.0171.471.471.4
Gemini-3-ProModality=Text2026.0142.928.6—
Gemini-3-ProModality=Image2026.0128.628.628.6
GPT-5Modality=Text2026.0116.70—
Gemini-2.5-ProModality=Image2026.01000
Gemini-2.5-ProModality=Text2026.0100—
Qwen3-VL-235BModality=Image2026.01000
Qwen3-VL-235BModality=Text2026.0100—
Qwen2.5-VL-72BModality=Image2026.01000
Qwen2.5-VL-72BModality=Text2026.0100—