AI Model Intelligence Platform
← Benchmarks

Best AI Models for Math Benchmark

Ranking of AI models by Math benchmark score.

RankModelProviderScoreDatasetVersionDate
1Claude 3.5 SonnetAnthropic88.7mmluofficial-20242024-06-20
2GPT-4oOpenAI88.7mmluofficial-20242024-05-13
3Llama 3.1 405BMeta88.6mmluofficial-20242024-07-23
4DeepSeek V3DeepSeek88.5mmlupaper-20242024-12-26
5Claude 3 OpusAnthropic86.8mmluofficial-20242024-03-04
6Qwen2.5 72BAlibaba86.1mmluofficial-20242024-09-19
7Llama 3.3 70BMeta86mmluofficial-20242024-12-06
8Mistral Large 2Mistral84mmluofficial-20242024-07-24
9GPT-4o miniOpenAI82mmluofficial-20242024-07-18
10Gemini 1.5 ProGoogle81.9mmlutechreport-20242024-05-14
11GLM-4Zhipu81.5mmluofficial-20242024-01-16
12DeepSeek R1DeepSeek79.8aimepaper-20252025-01-20
13Gemini 1.5 FlashGoogle78.9mmlutechreport-20242024-05-14
14Claude 3 HaikuAnthropic75.1mmluofficial-20242024-03-04

Source

Dataset
mmlu
Version
official-2024
Date
2024-06-20

Benchmark

Related Resources

LeaderboardRelated RankingBenchmarksBenchmarks
Featured Models
Claude 3.5 Sonnet (88.7) →GPT-4o (88.7) →Llama 3.1 405B (88.6) →
Use Cases
Math →