SciEval-Leaderboard / Multimodal Model Disciplinary Leaderboard.csv
naonaowyh's picture
initial leaderboard
fc80ff8 verified
Raw
History Blame Contribute Delete
1.44 kB
Model,Type,Parameters,Life Sci. ,Astronomy,Earth Sci. ,Chemistry,Mat. Sci. ,Physics,Overall
Claude 4.5 Sonnet,Close,,43.86 ,34.23 ,64.66 ,71.27 ,83.23 ,40.36 ,56.27
Claude4-1-Opus,Close,,42.49 ,39.87 ,67.47 ,71.94 ,83.23 ,38.69 ,57.28
GPT-4o,Close,,54.46 ,30.73 ,63.08 ,72.37 ,61.54 ,32.09 ,52.38
GPT-5,Close,,59.49 ,44.57 ,74.43 ,81.62 ,93.54 ,39.91 ,65.59
GPT-o3,Close,,61.57 ,42.82 ,74.15 ,81.77 ,93.85 ,38.14 ,65.38
Gemini-2.5-Flash,Close,,35.61 ,31.94 ,68.21 ,75.12 ,62.15 ,32.07 ,50.85
Gemini-2.5-Pro,Close,,34.85 ,43.40 ,70.52 ,78.29 ,83.23 ,34.96 ,57.54
Grok-2-vision-1212,Close,,43.44 ,33.51 ,58.80 ,63.04 ,49.08 ,28.21 ,46.01
Ling-flash-2.0,Open,100B,28.83 ,48.12 ,69.55 ,66.80 ,63.69 ,37.98 ,52.49
Seed1.6-vision,Close,,58.53 ,32.21 ,69.46 ,75.48 ,68.92 ,39.24 ,57.31
DeepSeek-R1,Open,685B,27.15 ,0.11 ,63.14 ,77.28 ,44.46 ,20.00 ,38.69
GLM-4.5V,Open,106B,54.11 ,0.32 ,52.68 ,71.38 ,57.08 ,13.43 ,41.50
InternS1,Open,241B,53.70 ,29.55 ,68.52 ,75.28 ,81.69 ,31.62 ,56.73
Kimi-k2,Open,1040B,41.10 ,35.57 ,77.77 ,77.04 ,71.38 ,38.59 ,56.91
Llama 4 Maverick,Open,400B,42.92 ,31.91 ,59.82 ,68.06 ,80.77 ,38.97 ,53.74
Qwen3-VL-235B-A22B,Open,235B,56.69 ,29.84 ,67.81 ,77.39 ,81.85 ,49.93 ,60.58
Qwen3-Max,Open,1000B,38.95 ,75.64 ,67.38 ,76.38 ,69.54 ,41.04 ,61.49
GPT-5.1,Close,,60.54 ,42.02 ,73.88 ,76.45 ,61.08 ,32.44 ,57.73
Gemini-3-Pro,Close,,49.20 ,42.23 ,72.49 ,83.08 ,91.23 ,45.19 ,63.90