Skip to content

Benchmark

U-MATH algebra accuracy

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

Models measured
30
Scores recorded
90
Sources
1
IQ weight
2.19

Score distribution

59.7%
Mean
72.2%
Median
31%
Std dev
10%
10th percentile
92%
90th percentile
0%
Min
0%99.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.19 for chat models, measured over 30 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Llama 3.1 70B Instructalgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH65.6%
52Gemini 1.5 Proalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH60%
53Gemini 2.0 Flash Experimentalalgebra, visual, gpt-4o-2024-08-06, thinking· U-MATH and mu-MATH60%
54Llama 3.1 8B Instructalgebra, text, gpt-4o-2024-08-06· U-MATH and mu-MATH60%
55Qwen2 VL 7b Instructalgebra, text, gpt-4o-2024-08-06· U-MATH and mu-MATH58.7%
56LFM-7balgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH56.7%
57Llama 3.2 11B Vision Instructalgebra, text, gpt-4o-2024-08-06· U-MATH and mu-MATH52%
58Llama 3.1 8B Instructalgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH50.6%
59Qwen2 VL 7b Instructalgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH50.6%
60o1algebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH50%
61Gemini 1.5 Flashalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH46.7%
62Llama 3.2 11B Vision Instructalgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH43.9%
63Phi 4algebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH43.3%
64O1 Minialgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH40%
65Pixtral 12Balgebra, text, gpt-4o-2024-08-06· U-MATH and mu-MATH40%
66Pixtral 12Balgebra, all, gpt-4o-2024-08-06· U-MATH and mu-MATH37.2%
67GPT 4oalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH33.3%
68QVQ 72B Previewalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH33.3%
69Claude 3.5 Sonnetalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH30%
70Qwen2-VL (72B) Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH26.7%
71Llama 3.2 90B Vision Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH23.3%
72Nvidia Llama 3.1 Nemotron 70B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH23.3%
73Pixtral 12Balgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH23.3%
74Qwen2.5 7B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH20%
75DeepSeek R1algebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH16.7%
76Qwen2.5 72B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH16.7%
77GPT-4o minialgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH13.3%
78Mistral Large Instruct 2411algebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH13.3%
79Qwen2.5 32B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH13.3%
80DeepSeek V3algebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH10%
81o3 Minialgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH10%
82Qwen2 VL 7b Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH10%
83LFM-7balgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH6.7%
84Llama 3.3 70B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH6.7%
85Qwen2.5 Math 72B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH6.7%
86Qwen2.5 Math 7B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH6.7%
87Llama 3.1 8B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH3.3%
88Llama 3.2 11B Vision Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH3.3%
89QwQ 32B Previewalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH3.3%
90Llama 3.1 70B Instructalgebra, visual, gpt-4o-2024-08-06· U-MATH and mu-MATH0%

Showing 51–90 of 90 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)