AIME leaderboard
Problems from the American Invitational Mathematics Examination, a qualifier for the US Math Olympiad. Gemini 3.1 Pro Preview leads with 98.13%.
- Models
- 80
- Leader
- Gemini 3.1 Pro Preview
- Source
- Vals AI
- As of
- Oct 24, 2026
Open weightsProprietaryBest for the price
Full leaderboard
| Rank | Model | Lab | Accuracy | Cost / test | Harness | Price in / out |
|---|---|---|---|---|---|---|
| 1 | Gemini 3.1 Pro Preview | 98.13% | $2 / $12 | — | $2 / $12 | |
| 2 | GPT-5.2 | OpenAI | 96.88% | $1.75 / $14 | — | $1.75 / $14 |
| 3 | Muse Spark | Meta | 96.88% | N/A | — | — |
| 4 | Gemini 3 Pro | 96.68% | $2 / $12 | — | — | |
| 5 | GPT-5.4 | OpenAI | 96.67% | $2.5 / $15 | — | $2.50 / $15 |
| 6 | Grok 4.20 | xAI | 96.46% | $2 / $6 | — | $1.25 / $2.50 |
| 7 | Claude Opus 4.7 | Anthropic | 96.25% | $5 / $25 | — | $5 / $25 |
| 8 | Gemini 3 Flash | 95.63% | $0.5 / $3 | — | $0.50 / $3 | |
| 9 | Claude Opus 4.6 | Anthropic | 95.63% | $5 / $25 | — | $5 / $25 |
| 10 | GPT-5.4 mini | OpenAI | 95.63% | $0.75 / $4.5 | — | $0.75 / $4.50 |
| 11 | Kimi K2.5 | Moonshot AI | 95.63% | $0.6 / $3 | — | $0.50 / $2.50 |
| 13 | Qwen3.6-Plus | Alibaba | 94.58% | $0.5 / $3 | — | $0.33 / $1.95 |
| 14 | GPT-5 | OpenAI | 93.37% | $1.25 / $10 | — | $1.25 / $10 |
| 15 | GPT-5.1 | OpenAI | 93.33% | $1.25 / $10 | — | $1.25 / $10 |
| 16 | GLM-4.7 | Z.ai | 93.33% | $0.6 / $2.2 | — | $0.54 / $1.98 |
| 17 | GLM-4.6 | Z.ai | 92.71% | $0.6 / $2.2 | — | $0.55 / $2.20 |
| 18 | gpt-oss-120b | OpenAI | 92.60% | $0.15 / $0.6 | — | $0.15 / $0.60 |
| 19 | Qwen3.5-Flash | Alibaba | 92.50% | $0.1 / $0.4 | — | $0.065 / $0.26 |
| 20 | Claude Sonnet 4.6 | Anthropic | 92.29% | $3 / $15 | — | $3 / $15 |
| 21 | Grok 4.1 Fast | xAI | 91.88% | $0.2 / $0.5 | — | — |
| 22 | GLM-5.1 | Z.ai | 91.88% | $1 / $3.2 | — | $1.40 / $4.40 |
| 23 | GLM-5 | Z.ai | 91.67% | $1 / $3.2 | — | $1 / $3.20 |
| 24 | GPT-5 mini | OpenAI | 91.46% | $0.25 / $2 | — | $0.25 / $2 |
| 26 | MiniMax-M2.7 | MiniMax | 91.04% | $0.3 / $1.2 | — | $0.30 / $1.20 |
| 27 | Grok 4 | xAI | 90.56% | $3 / $15 | — | — |
| 28 | GPT-5.4 nano | OpenAI | 88.75% | $0.2 / $1.25 | — | $0.20 / $1.25 |
| 29 | MiniMax-M2.5 | MiniMax | 88.75% | $0.3 / $1.2 | — | $0.30 / $1.20 |
| 30 | Claude Sonnet 4.5 | Anthropic | 88.19% | $3 / $15 | — | $3 / $15 |
| 31 | GLM-4.5 | Z.ai | 86.67% | $0.6 / $2.2 | — | $0.60 / $2.20 |
| 32 | o3-mini | OpenAI | 86.46% | $1.1 / $4.4 | — | $1.10 / $4.40 |
| 33 | Qwen 3.5 Plus | Alibaba | 86.04% | $0.4 / $2.4 | — | $0.30 / $1.80 |
| 34 | gpt-oss-20b | OpenAI | 86.04% | $0.07 / $0.3 | — | $0.018 / $0.09 |
| 35 | Gemini 2.5 Pro | 85.83% | $1.25 / $10 | — | $1.25 / $10 | |
| 36 | Kimi K2 Thinking | Moonshot AI | 85.42% | $0.6 / $2.5 | — | $0.60 / $2.50 |
| 37 | o3 | OpenAI | 85.28% | $2 / $8 | — | $2 / $8 |
| 40 | Qwen3-235B-A22B | Alibaba | 83.96% | $0.22 / $0.88 | — | $0.09 / $0.55 |
| 41 | o4-mini | OpenAI | 83.67% | $1.1 / $4.4 | — | $1.10 / $4.40 |
| 42 | Magistral Medium | Mistral AI | 83.54% | $2 / $5 | — | — |
| 43 | Gemini 3.1 Flash-Lite Preview | 83.33% | $0.25 / $1.5 | — | $0.25 / $1.50 | |
| 44 | Claude Haiku 4.5 | Anthropic | 82.71% | $1 / $5 | — | $1 / $5 |
| 45 | GPT-5 nano | OpenAI | 81.18% | $0.05 / $0.4 | — | $0.05 / $0.40 |
| 49 | MiniMax-M2.1 Preview | MiniMax | 77.92% | $0.3 / $1.2 | — | $0.30 / $1.20 |
| 50 | Claude Opus 4.5 | Anthropic | 76.88% | $5 / $25 | — | $5 / $25 |
| 52 | DeepSeek-R1 | DeepSeek | 73.96% | $1.35 / $5.4 | — | $0.50 / $2.15 |
| 53 | o1 | OpenAI | 71.46% | $15 / $60 | — | $15 / $60 |
| 54 | Grok 3 Mini Reasoning | xAI | 70.63% | $0.3 / $0.5 | — | — |
| 55 | DeepSeek-V3.2-Exp | DeepSeek | 64.79% | $0.56 / $1.68 | — | $0.28 / $0.42 |
| 56 | Kimi K2 Instruct | Moonshot AI | 62.71% | $1 / $3 | — | — |
| 57 | Qwen3-Max-Preview | Alibaba | 60.69% | $1.2 / $6 | — | — |
| 58 | Grok 3 Preview | xAI | 58.75% | $3 / $15 | — | — |
| 63 | GPT-4.1 mini | OpenAI | 49.38% | $0.4 / $1.6 | — | $0.40 / $1.60 |
| 65 | Claude Opus 4.1 | Anthropic | 44.24% | $15 / $75 | — | $15 / $75 |
| 67 | Mistral Medium | Mistral AI | 42.29% | $0.4 / $2 | — | — |
| 69 | Claude Opus 4 | Anthropic | 41.25% | $15 / $75 | — | — |
| 70 | GPT-4.1 | OpenAI | 39.58% | $2 / $8 | — | $2 / $8 |
| 71 | Claude Sonnet 4 | Anthropic | 38.54% | $3 / $15 | — | — |
| 72 | Grok 4 Fast | xAI | 33.33% | $0.2 / $0.5 | — | — |
| 73 | Gemini 2.0 Flash | 29.79% | $0.1 / $0.4 | — | — | |
| 74 | DeepSeek-V3 | DeepSeek | 27.50% | $0.9 / $0.9 | — | $0.29 / $1.14 |
| 75 | Grok 4.1 Fast Non-Reasoning | xAI | 27.08% | $0.2 / $0.5 | — | — |
| 76 | GPT-4.1 nano | OpenAI | 26.46% | $0.1 / $0.4 | — | $0.10 / $0.40 |
| 77 | Gemini 2.5 Flash Preview | 26.25% | $0.1 / $0.4 | — | — | |
| 78 | Llama 4 Maverick | Fireworks AI | 25.21% | $0.22 / $0.88 | — | $0.19 / $0.65 |
| 79 | Claude 3.7 Sonnet | Anthropic | 22.29% | $3 / $15 | — | — |
| 80 | Llama 4 Scout | Meta | 18.96% | $0.18 / $0.59 | — | $0.10 / $0.30 |
| 81 | Gemini 1.5 Pro | 18.75% | $1.25 / $5 | — | — | |
| 82 | Gemini 1.5 Flash | 17.29% | $0.075 / $0.3 | — | — | |
| 83 | Llama 3.3 Instruct Turbo | Meta | 16.04% | $0.88 / $0.88 | — | — |
| 84 | Grok 2 | xAI | 15.21% | $2 / $10 | — | — |
| 86 | Command A | Cohere | 13.33% | $2.5 / $10 | — | $0.30 / $1.50 |
| 87 | GPT-4o | OpenAI | 11.88% | $2.5 / $10 | — | $2.50 / $10 |
| 88 | GPT-4o mini | OpenAI | 11.46% | $0.15 / $0.6 | — | $0.15 / $0.60 |
| 89 | Claude 3.5 Sonnet | Anthropic | 10.00% | $3 / $15 | — | — |
| 90 | Llama 3.3 Nemotron Super | Meta | 9.38% | N/A | — | — |
| 91 | Mistral Large | Mistral AI | 9.17% | $2 / $6 | — | $0.50 / $1.50 |
| 92 | Mistral Small | Mistral AI | 5.63% | $0.2 / $0.6 | — | — |
| 93 | Magistral Small 1.2 | Mistral AI | 3.54% | $0.075 / $0.3 | — | — |
| 94 | Claude 3.5 Haiku | Anthropic | 3.33% | $0.8 / $4 | — | — |
| 95 | Jamba 1.6 Mini | AI21 Labs | 0.42% | $0.2 / $0.4 | — | — |
| 96 | Jamba 1.6 Large | AI21 Labs | 0.42% | $2 / $8 | — | — |
Gemini 4 Argon
Claude Opus 5.5
Muse Wonder
MiMo-V2.6-Pro
Qwen3.8-Max
Kimi K3
GLM-5.3
GPT-5.5
ERNIE 5.1
DeepSeek-V4.1-Flash-20260910
Grok 4.20 Beta
amazon-nova-experimental-chat
Dola Seed 2.0 Pro
NVIDIA-Nemotron-3-Ultra-550B-A55B
Gemma 4 31B
Step 5 Preview


