Skip to content

Benchmark

FrontierMath

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

Models measured
71
Scores recorded
109
Sources
2
IQ weight
1.73

Score distribution

19.5%
Mean
16.4%
Median
18%
Std dev
1.4%
10th percentile
40.3%
90th percentile
0%
Min
0%89%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.73 for chat models, measured over 71 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1GPT-5.6 Sol· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim89%
2GPT-5.6 Terra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim84.9%
3GPT-5.6 Luna· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim78.6%
4GPT-5.4 Proxhigh· Epoch AI Benchmarking Hub data50%
5GPT 5.4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim47.6%
6GPT 5.4xhigh· Epoch AI Benchmarking Hub data47.6%
7Claude Opus 4.8max· Epoch AI Benchmarking Hub data47.2%
8Claude Opus 4.7xhigh· Epoch AI Benchmarking Hub data43.8%
9Claude Opus 4.6max· Epoch AI Benchmarking Hub data40.7%
10GPT 5.2xhigh· Epoch AI Benchmarking Hub data40.7%
11GPT 5.2high· Epoch AI Benchmarking Hub data40.3%
12GPT 5.2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim40.3%
13Claude Opus 4.632k· Epoch AI Benchmarking Hub data40%
14Claude Opus 4.664k· Epoch AI Benchmarking Hub data39.7%
15GPT-5.5 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim39.6%
16Muse Spark· Epoch AI Benchmarking Hub data39%
17Gemini 3.5 Flashhigh· Epoch AI Benchmarking Hub data39%
18Kimi K2.6· Epoch AI Benchmarking Hub data39%
19Claude Opus 4.6· Epoch AI Benchmarking Hub data38.3%
20Gemini 3 Pro Preview· Epoch AI Benchmarking Hub data37.6%
21Gemini 3.1 Pro Preview· Epoch AI Benchmarking Hub data36.9%
22GPT 5.2medium· Epoch AI Benchmarking Hub data36.9%
23Gemini 3 Flash· Epoch AI Benchmarking Hub data35.6%
24GPT 5.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim35.4%
25GLM 5.1· Epoch AI Benchmarking Hub data33.4%
26GPT 5high· Epoch AI Benchmarking Hub data32.4%
27Claude Sonnet 4.616k· Epoch AI Benchmarking Hub data32.4%
28GPT 5.1high· Epoch AI Benchmarking Hub data31%
29GPT-5.4 minihigh· Epoch AI Benchmarking Hub data28.3%
30Kimi K2.5· Epoch AI Benchmarking Hub data27.9%
31GPT 5medium· Epoch AI Benchmarking Hub data27.2%
32GPT-5 minihigh· Epoch AI Benchmarking Hub data27.2%
33GPT 5.1medium· Epoch AI Benchmarking Hub data26.9%
34GPT 5.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim26.7%
35GPT-5.1-Instant· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim26.7%
36GPT 5.2low· Epoch AI Benchmarking Hub data26.6%
37GPT 5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim26.3%
38Qwen3.6 Plus· Epoch AI Benchmarking Hub data26.2%
39GPT 5.4 Nanohigh· Epoch AI Benchmarking Hub data25.9%
40o4 Minihigh· Epoch AI Benchmarking Hub data24.8%
41Qwen3.6 Max Preview· Epoch AI Benchmarking Hub data23.1%
42DeepSeek V3.2· Epoch AI Benchmarking Hub data22.1%
43GPT-5 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim22.1%
44Kimi K2 Thinking· Epoch AI Benchmarking Hub data21.4%
45Qwen3.5 Plus· Epoch AI Benchmarking Hub data21%
46Claude Opus 4.5· Epoch AI Benchmarking Hub data20.7%
47Claude Opus 4.532k· Epoch AI Benchmarking Hub data20.7%
48Claude Opus 4.516k· Epoch AI Benchmarking Hub data20.3%
49GPT-5 minimedium· Epoch AI Benchmarking Hub data20.3%
50Grok 4· Epoch AI Benchmarking Hub data19.7%

Showing 1–50 of 109 results.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · Data by LLM Stats