Skip to main content
Meo

AA Math Index — AI model leaderboard

AI models ranked by AA Math Index, an aggregated third-party benchmark from artificial_analysis. Higher is better. Cross-referenced against our first-party meo scores and Effective Value (𝕍).

Ranking 32 models across the full field · as of 2026-06-07.

#ModelLabAA Math Index
1OpenAI: GPT-5.2 Chatopenai99.0
2OpenAI: GPT-5 Codexopenai98.7
3OpenAI: GPT-5.1-Codexopenai95.7
4OpenAI: GPT-5openai94.3
5OpenAI: GPT-5.1openai94.0
6OpenAI: gpt-oss-120bopenai93.4
7OpenAI: GPT-5.1-Codex-Miniopenai91.7
8OpenAI: GPT-5 Miniopenai90.7
9OpenAI: o4 Miniopenai90.7
10OpenAI: gpt-oss-20bopenai89.3
11OpenAI: o3openai88.3
12Prime Intellect: INTELLECT-3prime-intellect88.0
13Google: Gemini 2.5 Progoogle87.7
14OpenAI: GPT-5 Nanoopenai83.7
15Xiaomi: MiMo-V2-Flashxiaomi67.7
16Google: Gemini 2.5 Flashgoogle60.3
17Google: Gemini 3 Flash Previewgoogle55.7
18Google: Gemini 2.5 Flash Lite Preview 09-2025google46.7
19OpenAI: GPT-4.1 Miniopenai46.3
20OpenAI: GPT-4.1openai34.7
21Reka Flash 3rekaai33.7
22OpenAI: GPT-4.1 Nanoopenai24.0
23Google: Gemma 3 27Bgoogle20.7
24Meta: Llama 4 Maverickmeta-llama19.3
25Google: Gemma 3 12Bgoogle18.3
26Microsoft: Phi 4microsoft18.0
27OpenAI: GPT-4o-miniopenai14.7
28Meta: Llama 4 Scoutmeta-llama14.0
29Cohere: Command Acohere13.0
30Google: Gemma 3 4Bgoogle12.7
31Microsoft: Phi 4 Mini Instructmicrosoft6.7
32OpenAI: GPT-4oopenai6.0

Artificial Analysis (artificialanalysis.ai). Redistribution requires an AA commercial license.

← All rankingsMethodology & 𝕍 →