Loading...
Loading...
AIME 2025
40
Total Models
76.84
Average Score
23.30 - 100.00
Score Range
1
Max Score Achievable
Top 10 Model Performance
Top 10 of 40
Models Using This Benchmark(40)
Google(8 models)
96.70%
Gemini 2.5 Pro Preview
google
88.00%
Gemini 2.5 Pro Preview
google
83.00%
Gemini 2.5 Flash Preview
google
78.00%
Gemini 2.5 Flash Preview
google
72.00%
63.10%
Gemini 2.0 Flash
google
27.50%
Gemini Diffusion
google
23.30%
OpenAI(8 models)
GPT 5-Pro
openai
100.00%
GPT-5
openai
99.60%
gpt-oss-20b
openai
98.70%
o3
openai
98.40%
gpt-oss-120b
openai
97.90%
o4 Mini
openai
92.70%
GPT-5 mini
openai
91.10%
GPT-5 nano
openai
85.20%
xAI(6 models)
Grok 4 Heavy
x-ai
100.00%
Grok 4
x-ai
98.80%
Grok 3 Beta
x-ai
93.30%
Grok 3 Mini Beta
x-ai
90.80%
Grok 3 Mini
x-ai
83.00%
Grok 3
x-ai
57.50%
Qwen(5 models)
Nvidia(3 models)
72.50%
58.40%
47.10%
LG(2 models)
85.30%
45.20%
Microsoft(2 models)
Phi 4 Reasoning Plus
microsoft
78.00%
Phi 4 Reasoning
microsoft
62.90%
Mistral(2 models)
Magistral Medium
mistral
64.90%
Magistral Small
mistral
62.80%
Anthropic(1 model)
Claude Opus 4.1
anthropic
78.00%
DeepSeek(1 model)
R1
deepseek
85.50%
MiniMax(1 model)
MiniMax M1
minimax
76.90%
Moonshot(1 model)
Kimi K2 Instruct
moonshotai
49.50%