자동화된 페이지입니다. AI가 자동으로 작성하고 번역했으며 소프트웨어가 출처와 대조해 검증했습니다. 사람 편집자는 없습니다. SI-IQ는 공개 벤치마크 결과로 만든 지수이며 IQ 검사가 아닙니다. 작동 방식
FrontierMath Tier 4: AI 모델 결과
FrontierMath Tier 4은(는) Epoch AI에서 실행하는 수학 벤치마크입니다. 당사 데이터에서 68개 모델이 발표된 결과를 가지고 있습니다.
SI-IQ 지수에 포함됩니다(적재량 k = 2.32, 우연 수준 0%).
- 평가 기관
- Epoch AI
- 벤치마크 홈페이지
- epoch.ai/frontiermath
결과
| 1 | GPT-6.1 Sol OpenAI | 100.0% | 148144–152 | 8 |
| 2 | GPT-6 Astra OpenAI | 97.6% | 150146–153 | 8 |
| 3 | Claude Opus 5.5 Anthropic | 95.0% | 144141–148 | 8 |
| 4 | Claude Fable 5 Anthropic | 90.2% | 138134–141 | 8 |
| 5 | GPT-6 Sol OpenAI | 90.0% | 140137–144 | 8 |
| 6 | Claude Fable 5.1 Anthropic | 87.8% | 140136–143 | 8 |
| 7 | GPT-5.6 Sol OpenAI | 82.9% | 140137–143 | 8 |
| 8 | Claude Sonnet 5.5 Anthropic | 80.5% | 140137–143 | 8 |
| 9 | GPT-5.5 Pro OpenAI | 78.0% | 139135–142 | 5 데이터 부족 |
| 10 | Gemini 4 Argon Google DeepMind | 78.0% | 140137–143 | 8 |
| 11 | AI Co-Mathematician Google DeepMind | 75.6% | 137132–142 | 1 데이터 부족 |
| 12 | Claude Opus 5 Anthropic | 73.2% | 136133–140 | 8 |
| 13 | GPT-5.5 OpenAI | 72.5% | 138135–141 | 9 |
| 14 | GPT-5.6 Terra OpenAI | 70.7% | 134131–137 | 8 |
| 15 | Muse Wonder Meta AI | 70.7% | 136133–139 | 4 데이터 부족 |
| 16 | GPT-5.6 Luna OpenAI | 61.0% | 130127–133 | 8 |
| 17 | GPT-5.4 Pro OpenAI | 58.5% | 135131–138 | 5 데이터 부족 |
| 18 | Claude Opus 4.8 Anthropic | 56.1% | 132129–135 | 8 |
| 19 | GPT-6 Luna OpenAI | 56.1% | 130127–133 | 8 |
| 20 | GPT-5.4 OpenAI | 49.0% | 131128–133 | 9 |
| 21 | Claude Haiku 5.5 Anthropic | 46.3% | 129126–132 | 8 |
| 22 | Muse Spark 1.3 Meta AI | 46.3% | 128125–131 | 5 데이터 부족 |
| 23 | Qwen 3.8 Max Alibaba | 46.3% | 130127–133 | 7 데이터 부족 |
| 24 | GPT-5.2 Pro OpenAI | 46.0% | 132128–135 | 3 데이터 부족 |
| 25 | Kimi K3오픈 Moonshot | 39.0% | 127124–130 | 8 |
| 26 | Gemini 3.7 Flash Google DeepMind | 36.6% | 131128–134 | 8 |
| 27 | Qwen3.7-Max Alibaba | 34.1% | 125122–128 | 9 |
| 28 | Qwen3.8 Max (0902) Alibaba | 34.1% | 127123–130 | 6 데이터 부족 |
| 29 | Claude Opus 4.7 Anthropic | 31.7% | 127124–130 | 9 |
| 30 | Grok 4.6 xAI | 31.7% | 126123–129 | 8 |
| 31 | Muse Spark Open오픈 Meta AI | 31.7% | 127124–130 | 6 |
| 32 | GPT-5.2 OpenAI | 31.7% | 125122–128 | 10 |
| 33 | Claude Sonnet 5 Anthropic | 29.3% | 126123–129 | 8 |
| 34 | GLM-5.2오픈 Z.ai (Zhipu AI) | 29.3% | 121118–124 | 9 |
| 35 | GLM-5.3오픈 Z.ai (Zhipu AI) | 29.3% | 125122–128 | 8 |
| 36 | Claude Opus 4.6 Anthropic | 26.8% | 123120–126 | 9 |
| 37 | DeepSeek V4 Pro 0813오픈 DeepSeek | 26.8% | 128125–131 | 8 |
| 38 | DeepSeek V4.1 Flash오픈 DeepSeek | 26.8% | 128125–131 | 8 |
| 39 | Gemini 3.1 Pro Google DeepMind | 26.8% | 127124–130 | 9 |
| 40 | Gemini 3.5 Flash Google DeepMind | 26.8% | 126123–129 | 9 |
| 41 | Mistral Large 4 Mistral AI | 26.8% | 122119–125 | 7 데이터 부족 |
| 42 | Kimi K2.6오픈 Moonshot | 25.6% | 121118–124 | 9 |
| 43 | DeepSeek V4 Flash 0731오픈 DeepSeek | 24.4% | 124121–127 | 8 |
| 44 | Grok 4.5 xAI | 24.4% | 124121–127 | 6 데이터 부족 |
| 45 | GPT-5 OpenAI | 22.0% | 121118–124 | 11 |
| 46 | Gemini 3.6 Flash Google DeepMind | 22.0% | 125122–128 | 8 |
| 47 | Gemini 3.8 Flash Google DeepMind | 22.0% | 130127–133 | 8 |
| 48 | GPT-5 Pro OpenAI | 19.5% | 124120–128 | 3 데이터 부족 |
| 49 | GLM-5.3-Flash오픈 Z.ai (Zhipu AI) | 17.1% | 120117–124 | 7 |
| 50 | Gemini 3 Flash Google DeepMind | 17.1% | 122119–125 | 9 |
| 51 | Grok 4.7 xAI | 17.1% | 123120–127 | 7 데이터 부족 |
| 52 | Inkling-Small오픈 Thinking Machines | 17.1% | 118115–122 | 7 |
| 53 | Grok 4.3 Beta xAI | 14.6% | 117114–121 | 6 데이터 부족 |
| 54 | GPT-5 mini OpenAI | 12.2% | 115112–118 | 11 |
| 55 | GPT-5.4 Nano OpenAI | 12.2% | 118115–121 | 8 |
| 56 | Kimi K2.7 Code오픈 Moonshot | 12.2% | 118115–122 | 7 |
| 57 | GPT-5.4 Mini OpenAI | 9.8% | 117113–120 | 8 |
| 58 | Claude Opus 4.5 Anthropic | 4.9% | 115111–118 | 10 |
| 59 | Inkling오픈 Thinking Machines | 4.9% | 114110–117 | 7 |
| 60 | o4-mini OpenAI | 4.9% | 112109–115 | 10 |
| 61 | Claude Opus 4.1 Anthropic | 2.4% | 105102–109 | 6 데이터 부족 |
| 62 | Claude Sonnet 4.5 Anthropic | 2.4% | 110107–113 | 11 |
| 63 | DeepSeek-V4-Pro오픈 DeepSeek | 2.4% | 117113–120 | 9 |
| 64 | GPT-5 nano OpenAI | 2.4% | 105102–109 | 10 |
| 65 | GPT-5.5 Instant OpenAI | 2.4% | 107103–111 | 4 데이터 부족 |
| 66 | Gemini 2.5 Pro (Jun 2025) Google DeepMind | 0.0% | 108104–112 | 7 |
| 67 | Gemini 3.5 Flash-Lite Google DeepMind | 0.0% | 109106–113 | 7 |
| 68 | o3-mini OpenAI | 0.0% | 105101–108 | 9 |
데이터 출처: Aider polyglot leaderboard data (Apache-2.0); Arena leaderboard dataset on Hugging Face (CC BY 4.0); Berkeley Function Calling Leaderboard, Gorilla project (Apache-2.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); API list prices via LiteLLM's price table (MIT); each price links to the developer's pricing page; Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..