FrontierMath Tiers 1-3 (2025 set)
Earlier FrontierMath set, superseded by v2. Counted as the same benchmark as v2, so a model with both results is not counted twice.
| Category | Math |
|---|---|
| Metric | Accuracy |
| Source | Epoch AI (evaluator) |
| License | CC BY 4.0 |
| Benchmark page | epoch.ai |
| Status | Scored |
| Weight | 1.3 within Math |
| Fit | Discrimination 1.91, noise 0.36 |
| Lineage | Shares weight with frontiermath |
Results
| # | Model | Provider | Result | Date |
|---|---|---|---|---|
| 1 | GPT-5.5 Pro High | OpenAI | 52.4% | 2026-04-23 |
| 2 | GPT-5.5 Extra High | OpenAI | 51.7% | 2026-04-23 |
| 3 | GPT-5.5 Pro Extra High | OpenAI | 51.0% | 2026-04-23 |
| 4 | GPT-5.4 Pro Extra High | OpenAI | 50.0% | 2026-03-06 |
| 5 | GPT-5.4 Extra High | OpenAI | 47.6% | 2026-03-06 |
| 6 | Claude Opus 4.8 Max | Anthropic | 47.2% | 2026-06-08 |
| 7 | Claude Opus 4.7 Extra High | Anthropic | 43.8% | 2026-04-17 |
| 8 | Claude Opus 4.6 Max | Anthropic | 40.7% | 2026-02-12 |
| 9 | GPT-5.2 Extra High | OpenAI | 40.7% | 2025-12-13 |
| 10 | GPT-5.2 High | OpenAI | 40.3% | 2025-12-11 |
| 11 | Claude Opus 4.6 Thinking | Anthropic | 40.0% | 2026-02-06 |
| 12 | Muse Spark | Meta | 39.0% | 2026-04-08 |
| 13 | Kimi K2.6 | Moonshot AI | 39.0% | 2026-05-07 |
| 14 | Gemini 3.5 Flash High | 39.0% | 2026-05-22 | |
| 15 | Claude Opus 4.6 | Anthropic | 38.3% | 2026-02-05 |
| 16 | Gemini 3 Pro Preview | 37.6% | 2025-11-21 | |
| 17 | Gemini 3.1 Pro Preview | 36.9% | 2026-02-19 | |
| 18 | GPT-5.2 Medium | OpenAI | 36.9% | 2025-12-11 |
| 19 | Gemini 3 Flash Preview | 35.6% | 2025-12-17 | |
| 20 | GLM-5.1 | Zhipu AI | 33.5% | 2026-05-11 |
| 21 | GPT-5 High | OpenAI | 32.4% | 2025-11-13 |
| 22 | Claude Sonnet 4.6 Thinking | Anthropic | 32.4% | 2026-02-20 |
| 23 | GPT-5.1 High | OpenAI | 31.0% | 2025-11-13 |
| 24 | Gemini 2.5 Deep Think | 29.0% | 2026-11-11 | |
| 25 | GPT-5.4 mini High | OpenAI | 28.3% | 2026-04-15 |
| 26 | Kimi K2.5 | Moonshot AI | 27.9% | 2026-02-03 |
| 27 | GPT-5 Mini High | OpenAI | 27.2% | 2025-11-13 |
| 28 | GPT-5 Medium | OpenAI | 27.2% | 2025-11-13 |
| 29 | GPT-5.1 Medium | OpenAI | 26.9% | 2025-11-17 |
| 30 | GPT-5.2 Low | OpenAI | 26.6% | 2025-12-11 |
| 31 | Qwen3.6 Plus | Alibaba | 26.2% | 2026-05-12 |
| 32 | GPT-5.4 nano High | OpenAI | 25.9% | 2026-04-15 |
| 33 | o4-mini High | OpenAI | 24.8% | 2025-11-13 |
| 34 | Qwen3.6 Max Preview | Alibaba | 23.1% | 2026-05-27 |
| 35 | DeepSeek V3.2 | DeepSeek | 22.1% | 2025-12-22 |
| 36 | Kimi K2 Thinking | Moonshot AI | 21.4% | 2025-12-05 |
| 37 | Qwen3.5 Plus | Alibaba | 21.0% | 2026-05-14 |
| 38 | Claude Opus 4.5 | Anthropic | 20.7% | 2025-11-25 |
| 39 | Claude Opus 4.5 Thinking | Anthropic | 20.7% | 2025-11-25 |
| 40 | GPT-5 Mini Medium | OpenAI | 20.3% | 2025-11-13 |
| 41 | Grok 4 | xAI | 19.7% | 2025-11-13 |
| 42 | o4-mini Medium | OpenAI | 19.0% | 2025-11-13 |
| 43 | o3 High | OpenAI | 18.7% | 2025-11-16 |
| 44 | GPT-5.1 Low | OpenAI | 17.3% | 2025-11-25 |
| 45 | o3 Medium | OpenAI | 16.9% | 2025-11-16 |
| 46 | GLM-5 | Zhipu AI | 16.4% | 2026-02-19 |
| 47 | Claude Sonnet 4.5 Thinking | Anthropic | 15.2% | 2025-11-16 |
| 48 | Gemini 2.5 Pro | 14.1% | 2025-11-24 | |
| 49 | o3-mini High | OpenAI | 12.4% | 2025-11-16 |
| 50 | o4-mini Low | OpenAI | 10.7% | 2025-11-16 |
| 51 | Gemini 2.5 Pro Preview 06-05 | 10.3% | 2025-07-03 | |
| 52 | Qwen3.6 Flash | Alibaba | 10.3% | 2026-05-12 |
| 53 | o3 Low | OpenAI | 9.7% | 2025-11-17 |
| 54 | Claude Sonnet 4.5 | Anthropic | 9.3% | 2025-11-16 |
| 55 | o1 High | OpenAI | 9.3% | 2025-03-07 |
| 56 | Qwen3 235B-A22B Thinking 2507 | Alibaba | 8.5% | 2025-12-11 |
| 57 | GPT-5 Nano High | OpenAI | 8.3% | 2025-10-30 |
| 58 | o3-mini Medium | OpenAI | 8.1% | 2025-03-06 |
| 59 | GPT-5 Nano Medium | OpenAI | 7.2% | 2025-08-07 |
| 60 | Claude Opus 4.1 Thinking | Anthropic | 7.2% | 2025-08-05 |
| 61 | Qwen3.5 Flash | Alibaba | 6.2% | 2026-05-12 |
| 62 | Claude Haiku 4.5 Thinking | Anthropic | 5.9% | 2025-10-22 |
| 63 | Claude Opus 4.1 | Anthropic | 5.9% | 2025-08-05 |
| 64 | Grok-3 mini High | xAI | 5.9% | 2025-04-10 |
| 65 | GPT-4.1 | OpenAI | 5.5% | 2025-04-14 |
| 66 | Gemini 2.5 Flash | 4.8% | 2025-12-18 | |
| 67 | Claude Opus 4 | Anthropic | 4.5% | 2025-07-04 |
| 68 | GPT-4.1 mini | OpenAI | 4.5% | 2025-04-14 |
| 69 | Claude Sonnet 4 | Anthropic | 4.1% | 2025-07-04 |
| 70 | Claude Haiku 4.5 | Anthropic | 4.1% | 2025-10-16 |
| 71 | Claude Sonnet 3.7 Thinking | Anthropic | 4.1% | 2025-03-06 |
| 72 | Claude Opus 4 Thinking | Anthropic | 4.1% | 2025-08-05 |
| 73 | GLM-4.6 | Zhipu AI | 3.8% | 2025-12-08 |
| 74 | Grok 3 | xAI | 3.8% | 2025-04-10 |
| 75 | Claude Sonnet 3.7 | Anthropic | 3.1% | 2025-03-06 |
| 76 | Grok-3 mini Low | xAI | 2.8% | 2025-04-10 |
| 77 | GLM-4.7 | Zhipu AI | 2.4% | 2026-01-30 |
| 78 | GPT-5.1 Non-reasoning | OpenAI | 2.1% | 2025-11-25 |
| 79 | Claude Sonnet 3.5 v2 | Anthropic | 2.1% | 2025-03-06 |
| 80 | DeepSeek-V3 | DeepSeek | 1.7% | 2025-03-07 |
| 81 | Gemini 2.0 Flash | 1.7% | 2025-03-09 | |
| 82 | Qwen Plus | Alibaba | 1.7% | 2025-05-14 |
| 83 | o1-mini Medium | OpenAI | 1.7% | 2025-03-06 |
| 84 | o1-mini High | OpenAI | 1.4% | 2025-03-06 |
| 85 | GPT-4.1 nano | OpenAI | 1.0% | 2025-04-14 |
| 86 | Qwen Max | Alibaba | 1.0% | 2025-04-02 |
| 87 | Claude 3.5 Sonnet | Anthropic | 1.0% | 2025-03-07 |
| 88 | Llama 4 Maverick | Meta | 0.7% | 2025-04-08 |
| 89 | Grok-2 (Dec 2024) | xAI | 0.7% | 2025-03-06 |
| 90 | Mistral Medium 3 | Mistral | 0.4% | 2025-05-07 |
| 91 | Mistral Large 2.1 | Mistral | 0.3% | 2025-03-06 |
| 92 | Claude Haiku 3.5 | Anthropic | 0.3% | 2025-03-07 |
| 93 | GPT-4o (2024-11-20) | OpenAI | 0.3% | 2025-03-06 |
| 94 | GPT-4o (2024-08-06) | OpenAI | 0.3% | 2025-03-07 |
| 95 | Llama 4 Scout | Meta | 0.0% | 2025-04-08 |
| 96 | Gemini 1.5 Flash (Sep 2024) | 0.0% | 2025-03-07 |