🔢 Use Case Rankings

Best LLM for Math (2026)

Top AI models for mathematical problem-solving, competition math, and quantitative reasoning.

#ModelCompositeAIMEMATHSpeedPrice
1
97.592%97.2%85 t/s$10/M
2
o3
OpenAI
97.191.6%96.7%58 t/s$40/M
3
Claude Opus 4
Anthropic
97.887.4%96.1%45 t/s$75/M
4
DeepSeek R1
DeepSeek
Open
92.379.8%97.3%32 t/s$2.19/M
5
o4-mini
OpenAI
91.579.4%93.2%145 t/s$4.4/M
6
88.855.3%88.1%55 t/s$15/M
7
86.755.2%87.4%203 t/s$0.6/M
8
Open
87.545.3%84.2%120 t/s$1.2/M
9
DeepSeek V3
DeepSeek
Open
86.139.2%90.2%68 t/s$1.1/M
10
90.340.5%85.6%79 t/s$15/M
Rankings are based on AIME, MATH benchmark results. See Methodology for scoring details.
Other Rankings

Browse More

⌨️ Best LLM for Coding🧠 Best LLM for Reasoning✍️ Best LLM for Writing🔬 Best LLM for Research📄 Best LLM for Long Context🔧 Best LLM for Tool Calling💬 Best LLM for Chat🌍 Best LLM for Multilingual