Skip to content

Benchmark

mu-MATH F1

Math benchmarks measure how well models solve maths problems. Scores are percentages. Higher is better.

Models measured
17
Scores recorded
85
Sources
1
IQ weight
1.50

Score distribution

74%
Mean
75%
Median
9.8%
Std dev
58.8%
10th percentile
85.8%
90th percentile
48.7%
Min
48.7%93.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.50 for chat models, measured over 17 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1o1Llama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH93.2%
2o1Gemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH90.6%
3O1 MiniLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH89.7%
4o1mu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH89.5%
5o1GPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH88.4%
6DeepSeek R1Llama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH88.2%
7QwQ 32B PreviewQwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH86.7%
8O1 MiniGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH86.2%
9Gemini 2.0 Flash ExperimentalGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct, thinking· U-MATH and mu-MATH85.8%
10O1 Minimu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH84.8%
11QwQ 32B PreviewLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH84%
12o1Qwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH83.8%
13Gemini 1.5 ProLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH83.6%
14Gemini 2.0 Flash ExperimentalLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct, thinking· U-MATH and mu-MATH83.3%
15QwQ 32B Previewmu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH83.2%
16DeepSeek R1mu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH82.2%
17DeepSeek R1Gemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH82%
18GPT 4oLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH81.8%
19O1 MiniGPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH81.2%
20Gemini 2.0 Flash Experimentalmu_math, F1, Qwen/Qwen2.5-72B-Instruct, thinking· U-MATH and mu-MATH81%
21Gemini 1.5 Promu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH80.7%
22Gemini 1.5 FlashLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH80.6%
23QwQ 32B PreviewGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH80%
24DeepSeek R1GPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH79.7%
25Gemini 1.5 ProGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH79.5%
26O1 MiniQwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH79.5%
27Qwen2.5 72B InstructLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH79.3%
28Mistral Large Instruct 2411Llama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH78.6%
29Gemini 1.5 ProGPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH78.2%
30QwQ 32B PreviewGPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH78%
31Claude 3.5 SonnetLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH77.9%
32Gemini 1.5 ProQwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH77.7%
33GPT 4oGPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH77.5%
34GPT 4omu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH77.4%
35Qwen2.5 Math 72B InstructLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH77.3%
36DeepSeek R1Qwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH76.8%
37Qwen2.5 Math 72B InstructGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH76.8%
38Mistral Large Instruct 2411mu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH76.6%
39GPT-4o miniLlama-3.1-70B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH76.2%
40Gemini 2.0 Flash ExperimentalQwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct, thinking· U-MATH and mu-MATH76%
41Mistral Large Instruct 2411GPT-4o, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH76%
42Qwen2.5 72B Instructmu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH75.6%
43Mistral Large Instruct 2411Gemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH75%
44Claude 3.5 Sonnetmu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH74.8%
45Gemini 1.5 Flashmu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH74.8%
46Gemini 2.0 Flash ExperimentalGPT-4o, F1, Qwen/Qwen2.5-72B-Instruct, thinking· U-MATH and mu-MATH74.3%
47GPT 4oQwen2.5-72B-Instruct, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH74.2%
48Qwen2.5 72B InstructGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH74.2%
49Qwen2.5 Math 72B Instructmu_math, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH74%
50Gemini 1.5 FlashGemini-1.5-Pro, F1, Qwen/Qwen2.5-72B-Instruct· U-MATH and mu-MATH73.9%

Showing 1–50 of 85 results.

Data credits: Toloka, U-MATH and mu-MATH public leaderboard. MIT Space; generated solution outputs retain underlying model terms and are not ingested. (licence)