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

Scientific Olympiad Reasoning on FrontierScience-Olympiad

43.5Biology Accuracy

GPT-5.2

Updated 20d ago

Evaluation Results

Method	Links
GPT-5.2 2026.02		43.5	89	74.3	77.1
o4-mini 2026.02		41.5	71.5	57.8	61.7
Gemini-3-Pro 2026.02		41	85.6	75.5	76.1
GPT-OSS-120B 2026.02		38.8	63.4	61.5	60
GPT-5 2026.02		35.5	81.5	67	69.7
GPT-5.1 2026.02		33.5	82.4	67.4	70
Grok-4 2026.02		33	73.2	67.2	66.2
P1-235B-A22B 2026.02		30	71	68	65.4
P1-VL-235B-A22B 2026.02		30	71.3	65.5	64.3
o3 2026.02		30	76.1	59	62.9
Nemotron-Cascade-2 2026.05		30	56.3	56	53.5
P1-30B-A3B 2026.05		27.5	57.5	57.5	54.5
P1-VL-235B-A22B 2026.02		26.3	77.2	67.3	67.1
DeepSeek-V3.2 2026.02		26.3	74.1	67.3	65.9
Qwen3-VL-235B-A22B 2026.02		26.3	61.9	57.8	56.3
Qwen3-235B-A22B 2026.02		26.3	58.1	57.3	54.5
Qwen3.6-35B-A3B 2026.05		25	74.4	65.5	65
SU-01 2026.05		25	69.4	62.5	61.5
Claude-Opus-4.5 2026.02		24	81.8	72.5	71.4
P1-235B-A22B 2026.02		22.5	67.2	65.8	62
Kimi-K2 2026.02		20	76.6	65	65.1
GLM-4.7 2026.02		20	70.6	69.5	65
P1-VL-30B-A3B 2026.02		20	58.8	54	52.5
o1 2026.02		20	50.9	40.3	42.5
Qwen3-VL-30B-A3B 2026.02		18.8	49.4	43.5	43.4
GLM-4.7-Flash 2026.05		17.5	60	54.5	53
Gemma-4-31B 2026.05		17.5	61.9	69	61
P1-30B-A3B 2026.02		15	61.9	56.3	54.4
Qwen3-30B-A3B 2026.02		10	47.8	45.3	42.8
GPT-4o 2026.02		3	12.4	14.1	12.3