Best AI models for math

Ranked by FrontierMath: original, unpublished problems written by professional mathematicians, from advanced to research level. Competition-style AIME results are shown alongside. Tested independently by Epoch AI.

Updated September 30, 2026

Our picks

#ModelFrontierMathiOverall indexiInputiOutputiContexti
1
93.7%
166.6$10$501.1M
2
91.2%
—$4$201M
3
90.2%
165.0$10$501M
4
89.1%
162.0$2$101.1M
5
88.8%
—$2$101M
6
87.7%
162.5$30$1801.1M
7
87.0%
163.6$10$501M
8
86.0%
159.3$2$121.1M
9
Claude Opus 5Anthropic
85.6%
162.7$5$251M
10
GPT-5.5OpenAI
85.3%
159.3$5$301.1M

per 1M tokens · USD

Quick answers

Which AI model is best at math?
GPT-6 Astra from OpenAI leads this ranking (FrontierMath: 93.7%). It costs $10 per million input tokens and $50 per million output tokens.
Is there a cheaper alternative?
GPT-5.6 Sol from OpenAI (FrontierMath: 89.1%) costs $2 / $10 per million tokens, about 5.0× less than GPT-6 Astra.

How we rank

Scores come from tests run independently by Epoch AI (CC BY 4.0), not from the companies’ own announcements. Prices come from OpenRouter’s public API. Both are refreshed automatically several times a day, so this ranking follows new releases and price changes.

Other rankings

CodingScienceCheapFreeOpen sourceLong contextImages & PDFsRankings →