OTIS Mock AIME
Competition math at AIME level.
Agent Index weight
0%
shown for context only
Models tested
202
run by Epoch AI on a shared harness
Results
| # | Model | Score | Normalised | Harness | Cost/task | Date |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5 Supported Anthropic | 100.0% | source | — | Aug 6, 2026 | |
| 2 | Claude Fable 5.1 Supported Anthropic | 100.0% | source | — | Sep 1, 2026 | |
| 3 | Claude Opus 5.5 Supported Anthropic | 100.0% | source | — | Sep 22, 2026 | |
| 4 | Claude Sonnet 5.5 Estimated Anthropic | 100.0% | source | — | Sep 29, 2026 | |
| 5 | Gemini 4 Argon Supported Google | 100.0% | source | — | Oct 13, 2026 | |
| 6 | Gemini 4 Carbon Supported Google | 100.0% | source | — | Nov 2, 2026 | |
| 7 | Gemini 4 Hydrogen Estimated Google | 100.0% | source | — | Nov 13, 2026 | |
| 8 | GPT-5.5 Supported OpenAI | 100.0% | source | — | Apr 24, 2026 | |
| 9 | GPT-5.5 Pro OpenAI | 100.0% | source | — | Apr 24, 2026 | |
| 10 | GPT-5.6 Sol Supported OpenAI | 100.0% | source | — | Jul 9, 2026 | |
| 11 | GPT-6.1 Sol Estimated OpenAI | 100.0% | source | — | Sep 29, 2026 | |
| 12 | GPT-6.1 Terra Estimated OpenAI | 100.0% | source | — | Oct 26, 2026 | |
| 13 | GPT-6.2 Astra OpenAI | 100.0% | source | — | Oct 27, 2026 | |
| 14 | GPT-6 Astra Supported OpenAI | 100.0% | source | — | Aug 30, 2026 | |
| 15 | GPT-6 Sol Estimated OpenAI | 100.0% | source | — | Sep 22, 2026 | |
| 16 | Kimi K3.1 Pro Moonshot AI | 100.0% | source | — | Nov 6, 2026 | |
| 17 | Qwen3.8 Max 0902 Alibaba (Qwen) | 100.0% | source | — | Sep 2, 2026 | |
| 18 | Claude Fable 5.2 Supported Anthropic | 99.7% | source | — | Nov 6, 2026 | |
| 19 | GPT-5.6 Terra Estimated OpenAI | 99.7% | source | — | Jul 9, 2026 | |
| 20 | DeepSeek V4.1 Flash 1017 Estimated DeepSeek | 99.4% | source | — | Oct 19, 2026 | |
| 21 | Mistral Large 4 Mistral | 99.4% | source | — | Nov 2, 2026 | |
| 22 | Muse Wonder Estimated Meta | 99.4% | source | — | Oct 12, 2026 | |
| 23 | Qwen3.8 Max Preview Estimated Alibaba (Qwen) | 99.4% | source | — | Aug 4, 2026 | |
| 24 | Doubao Seed Evolving (Nov 2026) ByteDance Seed | 99.2% | source | — | Nov 2, 2026 | |
| 25 | Grok 4.6 Supported xAI | 99.2% | source | — | Aug 14, 2026 | |
| 26 | Muse Spark 1.3 Estimated Meta | 99.2% | source | — | Sep 17, 2026 | |
| 27 | Claude Haiku 5.5 Estimated Anthropic | 98.9% | source | — | Oct 7, 2026 | |
| 28 | Claude Opus 5 Supported Anthropic | 98.9% | source | — | Jul 24, 2026 | |
| 29 | Gemini 3.8 Flash Supported Google | 98.9% | source | — | Sep 2, 2026 | |
| 30 | GPT-6 Luna OpenAI | 98.9% | source | — | Sep 22, 2026 | |
| 31 | DeepSeek V4 Pro 0813 DeepSeek | 98.6% | source | — | Aug 18, 2026 | |
| 32 | Claude Opus 4.8 Supported Anthropic | 98.3% | source | — | Jun 7, 2026 | |
| 33 | DeepSeek V4.1 Flash Estimated DeepSeek | 98.3% | source | — | Oct 8, 2026 | |
| 34 | GPT-5.6 Luna Estimated OpenAI | 98.3% | source | — | Jul 9, 2026 | |
| 35 | Grok 4.7 Estimated xAI | 98.1% | source | — | Sep 22, 2026 | |
| 36 | Muse Spark Open Meta | 98.1% | source | — | Oct 15, 2026 | |
| 37 | Claude Opus 4.7 Supported Anthropic | 97.8% | source | — | Apr 17, 2026 | |
| 38 | GPT-5.4 Supported OpenAI | 97.8% | source | — | Jul 15, 2026 | |
| 39 | Grok 4.5 Estimated xAI | 97.8% | source | — | Jul 8, 2026 | |
| 40 | Gemini 3.7 Flash Estimated Google | 97.2% | source | — | Aug 14, 2026 | |
| 41 | Kimi K3 Supported Moonshot AI | 97.2% | source | — | Jul 16, 2026 | |
| 42 | DeepSeek V4 Pro Estimated DeepSeek | 96.7% | source | — | Jun 17, 2026 | |
| 43 | Doubao Seed 2.1 Turbo ByteDance Seed | 96.1% | source | — | Oct 30, 2026 | |
| 44 | GPT-5.2 Supported OpenAI | 96.1% | source | — | Dec 11, 2025 | |
| 45 | Kimi K2.6 Estimated Moonshot AI | 96.1% | source | — | May 2, 2026 | |
| 46 | Gemini 3.1 Pro Preview Supported Google | 95.6% | source | — | Feb 20, 2026 | |
| 47 | Doubao Seed 2.1 Pro ByteDance Seed | 95.6% | source | — | Oct 30, 2026 | |
| 48 | Gemini 3.5 Flash Supported Google | 95.6% | source | — | May 25, 2026 | |
| 49 | Gemini 3 Flash Preview Supported Google | 95.6% | source | — | Aug 6, 2026 | |
| 50 | Kimi K2.7 Code Estimated Moonshot AI | 95.6% | source | — | Aug 7, 2026 | |
| 51 | Qwen3.7 Max Alibaba (Qwen) | 95.6% | source | — | Aug 7, 2026 | |
| 52 | Claude Sonnet 5 Supported Anthropic | 94.7% | source | — | Jul 1, 2026 | |
| 53 | Claude Opus 4.6 Supported Anthropic | 94.4% | source | — | Feb 6, 2026 | |
| 54 | DeepSeek V4 Flash 0731 DeepSeek | 94.4% | source | — | Aug 2, 2026 | |
| 55 | Gemini 3.6 Flash Estimated Google | 94.2% | source | — | Aug 2, 2026 | |
| 56 | Doubao-Seed-2.0 Pro ByteDance Seed | 93.9% | source | — | Oct 30, 2026 | |
| 57 | GLM-5.3-Flash Supported Z.ai (Zhipu) | 93.9% | source | — | Aug 26, 2026 | |
| 58 | GLM-5.1 Estimated Z.ai (Zhipu) | 93.3% | source | — | Aug 10, 2026 | |
| 59 | Grok 4.3 Beta xAI | 93.3% | source | — | Jun 17, 2026 | |
| 60 | Qwen3.6 Plus Estimated Alibaba (Qwen) | 93.3% | source | — | Aug 7, 2026 | |
| 61 | Qwen3.7 Plus Alibaba (Qwen) | 93.3% | source | — | Aug 7, 2026 | |
| 62 | Grok 4.20 (Non-Reasoning) Estimated xAI | 92.2% | source | — | Jul 13, 2026 | |
| 63 | Kimi K2.5 Supported Moonshot AI | 92.2% | source | — | Feb 2, 2026 | |
| 64 | Gemini 3 Pro Preview Supported Google | 91.4% | source | — | Nov 19, 2025 | |
| 65 | GPT-5 Supported OpenAI | 91.4% | source | — | Oct 29, 2025 | |
| 66 | Doubao Seed 2.0 Lite (Apr 2026) ByteDance Seed | 91.1% | source | — | Oct 30, 2026 | |
| 67 | GLM-5.3 Supported Z.ai (Zhipu) | 91.1% | source | — | Aug 24, 2026 | |
| 68 | Qwen3.6 27B Alibaba (Qwen) | 91.1% | source | — | Aug 7, 2026 | |
| 69 | Qwen3.6 Max Preview Estimated Alibaba (Qwen) | 91.1% | source | — | Aug 7, 2026 | |
| 70 | Inkling Small thinkingmachines | 90.0% | source | — | Aug 14, 2026 | |
| 71 | Muse Spark Meta | 88.9% | source | — | Apr 8, 2026 | |
| 72 | GPT-5.4 mini OpenAI | 88.9% | source | — | Aug 7, 2026 | |
| 73 | GPT OSS 120B Supported OpenAI | 88.9% | source | — | Dec 11, 2025 | |
| 74 | Inkling Estimated thinkingmachines | 88.9% | source | — | Aug 5, 2026 | |
| 75 | Qwen3.5 397B-A17B Alibaba (Qwen) | 88.9% | source | — | Aug 7, 2026 | |
| 76 | GPT-5.1 Supported OpenAI | 88.6% | source | — | Nov 13, 2025 | |
| 77 | DeepSeek V3.2 Supported DeepSeek | 87.8% | source | — | Dec 16, 2025 | |
| 78 | GPT-5.4 nano OpenAI | 87.8% | source | — | Apr 14, 2026 | |
| 79 | GPT-5 Mini Estimated OpenAI | 86.7% | source | — | Oct 30, 2025 | |
| 80 | Nemotron 3 Ultra NVIDIA | 86.7% | source | — | Aug 10, 2026 | |
| 81 | Qwen3 235B A22B Thinking 2507 Alibaba (Qwen) | 86.7% | source | — | Dec 10, 2025 | |
| 82 | Qwen3.5 Plus Estimated Alibaba (Qwen) | 86.7% | source | — | Aug 7, 2026 | |
| 83 | Qwen3.6 35B-A3B Alibaba (Qwen) | 86.7% | source | — | Aug 7, 2026 | |
| 84 | Qwen3.7 Flash Alibaba (Qwen) | 86.7% | source | — | Aug 7, 2026 | |
| 85 | GLM-5.2 Supported Z.ai (Zhipu) | 86.4% | source | — | Jun 25, 2026 | |
| 86 | Claude Opus 4.5 (latest) Supported Anthropic | 86.1% | source | — | Nov 24, 2025 | |
| 87 | Claude Sonnet 4.6 Supported Anthropic | 85.8% | source | — | Feb 20, 2026 | |
| 88 | Gemini 2.5 Pro Supported Google | 84.7% | source | — | Nov 16, 2025 | |
| 89 | o3 Supported OpenAI | 84.4% | source | — | Aug 7, 2026 | |
| 90 | Qwen3.5 Flash Alibaba (Qwen) | 84.4% | source | — | Aug 7, 2026 | |
| 91 | Qwen3.6 Flash Alibaba (Qwen) | 84.4% | source | — | Aug 7, 2026 | |
| 92 | Grok 4 Supported xAI | 84.0% | source | — | — | |
| 93 | GLM-4.7 Estimated Z.ai (Zhipu) | 83.3% | source | — | Jan 29, 2026 | |
| 94 | Kimi K2 Thinking Estimated Moonshot AI | 83.1% | source | — | Nov 13, 2025 | |
| 95 | Gemma 4 26B A4B Google | 82.2% | source | — | Aug 28, 2026 | |
| 96 | o4-mini Estimated OpenAI | 81.7% | source | — | Apr 16, 2025 | |
| 97 | GPT-5 Nano OpenAI | 81.1% | source | — | Oct 31, 2025 | |
| 98 | Gemini 3.1 Flash Lite Google | 80.0% | source | — | Aug 6, 2026 | |
| 99 | GLM-5 Supported Z.ai (Zhipu) | 80.0% | source | — | Feb 12, 2026 | |
| 100 | Claude Sonnet 4.5 (latest) Supported Anthropic | 77.8% | source | — | Oct 28, 2025 | |
| 101 | Grok-3 mini xAI | 77.8% | source | — | Apr 10, 2025 | |
| 102 | o3-mini Estimated OpenAI | 76.9% | source | — | Feb 27, 2025 | |
| 103 | Gemma 4 31B IT Google | 73.3% | source | — | Aug 6, 2026 | |
| 104 | o1 Estimated OpenAI | 73.3% | source | — | Feb 27, 2025 | |
| 105 | Qwen3 Max Estimated Alibaba (Qwen) | 73.3% | source | — | Oct 6, 2025 | |
| 106 | Gemini 2.5 Flash Estimated Google | 73.1% | source | — | Apr 22, 2025 | |
| 107 | Claude Sonnet 4 Supported Anthropic | 71.1% | source | — | May 22, 2025 | |
| 108 | Gemini 3.5 Flash Lite Google | 71.1% | source | — | Aug 6, 2026 | |
| 109 | MiniMax-M3 Estimated MiniMax | 71.1% | source | — | Aug 10, 2026 | |
| 110 | Qwen3-30B-A3B-Thinking (Jul 2025) Alibaba (Qwen) | 70.3% | source | — | Aug 30, 2026 | |
| 111 | Qwen3.5 35B-A3B Alibaba (Qwen) | 70.0% | source | — | Aug 30, 2026 | |
| 112 | Doubao-Seed 1.6 ByteDance Seed | 69.4% | source | — | Oct 30, 2026 | |
| 113 | Claude Opus 4.1 (latest) Supported Anthropic | 68.9% | source | — | Aug 5, 2025 | |
| 114 | GPT Chat Latest OpenAI | 68.1% | source | — | Aug 2, 2026 | |
| 115 | seed-oss-36b-instruct Unknown | 67.5% | source | — | Aug 28, 2026 | |
| 116 | Qwen3 32B Alibaba (Qwen) | 66.9% | source | — | Aug 30, 2026 | |
| 117 | Claude Haiku 4.5 Estimated Anthropic | 66.7% | source | — | Oct 22, 2025 | |
| 118 | DeepSeek-R1 Estimated DeepSeek | 66.4% | source | — | May 29, 2025 | |
| 119 | Qwen3 14B Alibaba (Qwen) | 66.4% | source | — | Aug 28, 2026 | |
| 120 | GPT OSS 20B OpenAI | 65.3% | source | — | Aug 27, 2026 | |
| 121 | Claude-Opus-4 Supported Anthropic | 64.4% | source | — | May 28, 2025 | |
| 122 | Qwen3 30B A3B Alibaba (Qwen) | 62.8% | source | — | Aug 28, 2026 | |
| 123 | Qwen3 30B A3B Instruct 2507 Alibaba (Qwen) | 62.2% | source | — | Aug 30, 2026 | |
| 124 | Qwen3.5 9B Alibaba (Qwen) | 61.7% | source | — | Aug 27, 2026 | |
| 125 | Qwq 32B Alibaba (Qwen) | 59.2% | source | — | Aug 28, 2026 | |
| 126 | GLM-4.7-Flash Z.ai (Zhipu) | 58.3% | source | — | Aug 30, 2026 | |
| 127 | Claude 3.7 Sonnet Supported Anthropic | 57.8% | source | — | Mar 13, 2025 | |
| 128 | Gemini 2.0 Flash Thinking (Jan 2025) Google | 57.8% | source | — | Mar 7, 2025 | |
| 129 | Qwen3 8B Alibaba (Qwen) | 56.1% | source | — | Aug 27, 2026 | |
| 130 | Qwen3.5 4B Alibaba (Qwen) | 55.8% | source | — | Aug 28, 2026 | |
| 131 | DeepSeek-R1-Distill-Qwen-32B DeepSeek | 55.6% | source | — | Aug 28, 2026 | |
| 132 | Grok 3 xAI | 55.6% | source | — | Apr 10, 2025 | |
| 133 | qwen3-4b-instruct-2507 Unknown | 52.2% | source | — | Aug 28, 2026 | |
| 134 | DeepSeek-R1-Distill-Llama-70B DeepSeek | 51.4% | source | — | Mar 7, 2025 | |
| 135 | DeepSeek-R1-Distill-Qwen-14B DeepSeek | 50.6% | source | — | Aug 28, 2026 | |
| 136 | o1-mini OpenAI | 46.9% | source | — | Mar 6, 2025 | |
| 137 | GPT-4.1 mini OpenAI | 44.7% | source | — | Apr 14, 2025 | |
| 138 | DeepSeek R1 0528 Qwen3 8B DeepSeek | 43.9% | source | — | Aug 28, 2026 | |
| 139 | GPT-4.1 OpenAI | 38.3% | source | — | Apr 14, 2025 | |
| 140 | DeepSeek V3 0324 DeepSeek | 37.8% | source | — | Apr 1, 2025 | |
| 141 | GPT-4.5 OpenAI | 37.8% | source | — | Feb 28, 2025 | |
| 142 | Mistral Medium 3 Mistral | 32.2% | source | — | May 7, 2025 | |
| 143 | Gemini-2.0-Flash Google | 31.1% | source | — | Feb 25, 2025 | |
| 144 | Mistral Small 3.2 Mistral | 30.3% | source | — | Aug 30, 2026 | |
| 145 | Magistral Small 1.0 Mistral | 30.0% | source | — | Aug 6, 2026 | |
| 146 | GPT-4.1 nano OpenAI | 28.9% | source | — | Apr 14, 2025 | |
| 147 | Magistral Small 1.2 Mistral | 28.1% | source | — | Aug 30, 2026 | |
| 148 | Gemini 1.5 Pro (Feb 2024) Google | 23.1% | source | — | Feb 25, 2025 | |
| 149 | Gemma 3 27B Google | 22.5% | source | — | Aug 27, 2026 | |
| 150 | DeepSeek-R1-Distill-Qwen-1.5B DeepSeek | 21.4% | source | — | Aug 28, 2026 | |
| 151 | Llama 4 Maverick Meta | 20.6% | source | — | Apr 8, 2025 | |
| 152 | Qwen Plus Alibaba (Qwen) | 17.8% | source | — | Apr 7, 2025 | |
| 153 | Gemma 3 12B Google | 16.7% | source | — | Aug 28, 2026 | |
| 154 | Gemini 1.5 Flash (Sep 2024) Google | 16.3% | source | — | Feb 25, 2025 | |
| 155 | Qwen2.5-Max Alibaba (Qwen) | 16.1% | source | — | Apr 1, 2025 | |
| 156 | DeepSeek-V3 DeepSeek | 15.8% | source | — | Feb 25, 2025 | |
| 157 | Phi 4 Microsoft | 13.8% | source | — | Feb 25, 2025 | |
| 158 | Grok-2 (Dec 2024) xAI | 11.5% | source | — | Feb 25, 2025 | |
| 159 | Llama 3.1-405B Meta | 9.7% | source | — | Feb 25, 2025 | |
| 160 | Claude-Sonnet-3.5 Estimated Anthropic | 8.5% | source | — | Feb 25, 2025 | |
| 161 | Mistral Large 2 (Nov 2024) Mistral | 8.5% | source | — | Feb 25, 2025 | |
| 162 | Qwen2.5-72B Estimated Alibaba (Qwen) | 8.1% | source | — | Feb 25, 2025 | |
| 163 | Qwen3-1.7B Alibaba (Qwen) | 8.1% | source | — | Aug 28, 2026 | |
| 164 | Llama 4 Scout Meta | 7.8% | source | — | Apr 8, 2025 | |
| 165 | Gemma 3 4B Google | 7.5% | source | — | Aug 28, 2026 | |
| 166 | Qwen2.5-32B Alibaba (Qwen) | 7.4% | source | — | Feb 25, 2025 | |
| 167 | GPT-4o mini OpenAI | 6.9% | source | — | Jul 30, 2025 | |
| 168 | GPT-4 Turbo OpenAI | 6.7% | source | — | Feb 27, 2025 | |
| 169 | Mistral Small 3 Mistral | 6.7% | source | — | Aug 30, 2026 | |
| 170 | GPT-4o (2024-08-06) Supported OpenAI | 6.4% | source | — | Feb 25, 2025 | |
| 171 | Qwen Turbo Alibaba (Qwen) | 6.1% | source | — | Apr 7, 2025 | |
| 172 | Mistral Small 3.1 Mistral | 5.8% | source | — | Mar 18, 2025 | |
| 173 | Llama 3.3 70B Meta | 5.1% | source | — | Feb 25, 2025 | |
| 174 | Claude 3 Opus Estimated Anthropic | 4.7% | source | — | Feb 25, 2025 | |
| 175 | Gemini 1.5 Flash 8B Google | 4.6% | source | — | Feb 25, 2025 | |
| 176 | Tulu 3 (Tülu 3) 70B Allen Institute for AI | 4.4% | source | — | Mar 7, 2025 | |
| 177 | Claude 3.5 Haiku Anthropic | 4.3% | source | — | Feb 25, 2025 | |
| 178 | Llama 3-70B Meta | 4.3% | source | — | Feb 25, 2025 | |
| 179 | Llama 3.1-70B Meta | 3.6% | source | — | Feb 25, 2025 | |
| 180 | granite-4.0-micro Unknown | 2.8% | source | — | Aug 30, 2026 | |
| 181 | Llama 3.2 90B Meta | 2.6% | source | — | Feb 25, 2025 | |
| 182 | Claude 2 Anthropic | 2.5% | source | — | Mar 7, 2025 | |
| 183 | Claude 3 Sonnet Anthropic | 2.5% | source | — | Feb 25, 2025 | |
| 184 | Hermes 2 Theta Llama-3 70B Nous Research | 2.5% | source | — | Apr 1, 2025 | |
| 185 | Qwen2.5-7B Alibaba (Qwen) | 2.5% | source | — | Aug 30, 2026 | |
| 186 | GPT-3.5-turbo OpenAI | 2.2% | source | — | Aug 7, 2026 | |
| 187 | Claude 2.1 Anthropic | 1.9% | source | — | Mar 7, 2025 | |
| 188 | Llama 3-8B Meta | 1.9% | source | — | Aug 30, 2026 | |
| 189 | Mistral Large Mistral | 1.9% | source | — | Feb 25, 2025 | |
| 190 | Claude-Haiku-3 Anthropic | 1.8% | source | — | Feb 25, 2025 | |
| 191 | Llama 3.1-8B Meta | 1.7% | source | — | Aug 27, 2026 | |
| 192 | Gemma 2 27B Google | 1.4% | source | — | Mar 7, 2025 | |
| 193 | Gemini 1.0 Pro Google | 1.1% | source | — | Feb 25, 2025 | |
| 194 | Gemma 3 1B Google | 1.1% | source | — | Aug 28, 2026 | |
| 195 | GPT-4 OpenAI | 1.1% | source | — | Oct 23, 2025 | |
| 196 | DeepSeek LLM 67B DeepSeek | 0.8% | source | — | Aug 28, 2026 | |
| 197 | granite-4.0-1b Unknown | 0.8% | source | — | Aug 30, 2026 | |
| 198 | Gemma 2 9B Google | 0.6% | source | — | Mar 7, 2025 | |
| 199 | Llama 3.2 1B Instruct Meta | 0.6% | source | — | Aug 30, 2026 | |
| 200 | granite-4.0-350m Unknown | 0.3% | source | — | Aug 28, 2026 | |
| 201 | Mistral 7B v0.3 Mistral | 0.3% | source | — | Aug 30, 2026 | |
| 202 | Llama 2-70B Meta | 0.0% | source | — | Feb 25, 2025 |