Benchmarks / External Composite Indices
AA Math Index
Artificial Analysis's math composite. Displayed for reference only.
maintained by Artificial Analysis · 109 observations on record
Versions
1 version
| Version | Tasks |
|---|---|
| AA Math Index current aa-math-index | — |
Current leaderboard
AA Math Index · Index points · top 25 of 109 systems
| # | System | Index points |
|---|---|---|
| 1 | GPT 5.2 (2025-12-11)xhigh effort OpenAI | 99.0 points |
| 2 | GPT 5 Codexhigh effort OpenAI | 98.7 points |
| 3 | Gemini 3 Flash Previewthinking Google DeepMind | 97.0 points |
| 4 | GPT 5.2 (2025-12-11)medium effort OpenAI | 96.7 points |
| 5 | Mimo V2 Flashthinking xiaomi | 96.3 points |
| 6 | GPT 5.1 Codexhigh effort Unknown | 95.7 points |
| 7 | Gemini 3 Pro Previewhigh effort Google DeepMind | 95.7 points |
| 8 | GLM 4.7thinking Z.ai | 95.0 points |
| 9 | Kimi K2 Thinking Unknown | 94.7 points |
| 10 | GPT 5 (2025-08-07)high effort OpenAI | 94.3 points |
| 11 | GPT 5.1 (2025-11-13)high effort OpenAI | 94.0 points |
| 12 | GPT OSS 120Bhigh effort OpenAI | 93.4 points |
| 13 | Grok 4 xAI | 92.7 points |
| 14 | Deepseek V3 2thinking DeepSeek | 92.0 points |
| 15 | GPT 5.1 Codex Minihigh effort OpenAI | 91.7 points |
| 16 | GPT 5 (2025-08-07)medium effort OpenAI | 91.7 points |
| 17 | Claude Opus 4.5thinking Anthropic | 91.3 points |
| 18 | Qwen3 235B A22b 2507thinking Alibaba | 91.0 points |
| 19 | GPT 5 Mini (2025-08-07)high effort OpenAI | 90.7 points |
| 20 | O4 Mini (2025-04-16)high effort OpenAI | 90.7 points |
| 21 | Grok 4 Fastthinking xAI | 89.7 points |
| 22 | Deepseek V3 1 Terminusthinking DeepSeek | 89.7 points |
| 23 | Deepseek V3 1thinking DeepSeek | 89.7 points |
| 24 | GPT OSS 20Bhigh effort OpenAI | 89.3 points |
| 25 | O3 (2025-04-16) OpenAI | 88.3 points |
One row per evaluated system — reasoning-effort variants rank separately.