| # | Model | Score | Relative | Setting | Released | Run | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | GPT-6 Astra OpenAI | 95.8% ±1.4 | max | 3 Sep 2026 | 30 Aug 2026 | Source ↗ | |
| 2 | Gemini 3.8 Flash Google | 95.4% ±1.4 | high | 2 Sep 2026 | 2 Sep 2026 | Source ↗ | |
| 3 | Gemini 3.7 Flash Google | 94.8% ±1.3 | high | 13 Aug 2026 | 14 Aug 2026 | Source ↗ | |
| 4 | GPT-5.4 Pro (xhigh) OpenAI | 94.6% ±1.6 | xhigh | 5 Mar 2026 | 20 Mar 2026 | Source ↗ | |
| 5 | Gemini 3.1 Pro Preview Google | 94.4% ±1.6 | high | 19 Feb 2026 | 6 Aug 2026 | Log ↗ | |
| 6 | Gemini 3.6 Flash Google | 94.1% ±1.4 | high | 21 Jul 2026 | 2 Aug 2026 | Source ↗ | |
| 7 | Grok 4.6 SpaceXAI | 94.0% ±1.4 | high | 12 Aug 2026 | 12 Aug 2026 | Source ↗ | |
| 8 | GPT-5.5 (xhigh) OpenAI | 94.0% ±1.5 | xhigh | 23 Apr 2026 | 24 Apr 2026 | Log ↗ | |
| 9 | GPT-5.5 Pro (xhigh) OpenAI | 93.9% ±1.6 | xhigh | 23 Apr 2026 | 24 Apr 2026 | Log ↗ | |
| 10 | Claude Opus 5 Anthropic | 93.9% ±1.5 | max | 24 Jul 2026 | 24 Jul 2026 | Source ↗ | |
| 11 | GPT-5.6 Sol OpenAI | 93.5% ±1.6 | max | 9 Jul 2026 | 9 Jul 2026 | Source ↗ | |
| 12 | Grok 4.5 SpaceXAI | 93.4% ±1.4 | high | 8 Jul 2026 | 8 Jul 2026 | Source ↗ | |
| 13 | GPT-5.6 Terra OpenAI | 93.3% ±1.5 | max | 9 Jul 2026 | 9 Jul 2026 | Source ↗ | |
| 14 | GPT-5.4 (xhigh) OpenAI | 93.3% ±1.8 | xhigh | 5 Mar 2026 | 6 Mar 2026 | Log ↗ | |
| 15 | Grok 4.6 (xhigh) xAI | 93.2% ±1.5 | xhigh | 12 Aug 2026 | 14 Aug 2026 | Source ↗ | |
| 16 | Kimi K3 Open source MoonshotAI | 93.1% ±1.5 | max | 16 Jul 2026 | 16 Jul 2026 | Source ↗ | |
| 17 | Gemini 3.5 Flash Google | 92.8% ±1.6 | high | 19 May 2026 | 22 May 2026 | Log ↗ | |
| 18 | Qwen3.8 Max (xhigh) Alibaba | 92.7% ±1.7 | xhigh | 2 Aug 2026 | 4 Aug 2026 | Source ↗ | |
| 19 | Gemini 3 Pro Preview Google DeepMind | 92.6% ±1.7 | — | 18 Nov 2025 | 19 Nov 2025 | Log ↗ | |
| 20 | Qwen3.8 Max (0902) (xhigh) Alibaba | 92.3% ±1.7 | xhigh | 1 Sep 2026 | 2 Sep 2026 | Source ↗ | |
| 21 | GLM 5.2 Open source Z.ai | 91.9% ±1.6 | max | 16 Jun 2026 | 24 Jun 2026 | Source ↗ | |
| 22 | DeepSeek V4 Pro 0813 Open source DeepSeek | 91.7% ±1.5 | max | 13 Aug 2026 | 18 Aug 2026 | Source ↗ | |
| 23 | GPT-5.6 Luna OpenAI | 91.6% ±1.7 | max | 9 Jul 2026 | 9 Jul 2026 | Source ↗ | |
| 24 | GPT-5.2 (xhigh) OpenAI | 91.4% ±1.8 | xhigh | 11 Dec 2025 | 13 Dec 2025 | Log ↗ | |
| 25 | DeepSeek V4 Flash 0731 Open source DeepSeek | 91.0% ±1.8 | max | 31 Jul 2026 | 2 Aug 2026 | Source ↗ | |
| 26 | Claude Opus 4.8 Anthropic | 91.0% ±1.9 | max | 28 May 2026 | 7 Jun 2026 | Source ↗ | |
| 27 | GLM 5.3 Open source Z.ai | 90.9% ±1.6 | max | 14 Aug 2026 | 24 Aug 2026 | Source ↗ | |
| 28 | MiniMax M3 Open source MiniMax | 90.9% ±2.0 | — | 1 Jun 2026 | 10 Aug 2026 | Log ↗ | |
| 29 | Qwen3.7 Max Qwen | 90.9% ±2.0 | — | 19 May 2026 | 7 Aug 2026 | Log ↗ | |
| 30 | DeepSeek v4 Pro (high) Open source DeepSeek | 90.9% ±2.0 | high | 24 Apr 2026 | 6 Aug 2026 | Log ↗ | |
| 31 | Kimi K2.6 Open source MoonshotAI | 90.8% ±1.7 | — | 20 Apr 2026 | 1 May 2026 | Log ↗ | |
| 32 | GPT-5.5 (low) OpenAI | 90.7% ±1.8 | low | 23 Apr 2026 | 5 May 2026 | Log ↗ | |
| 33 | Claude Sonnet 5 (xhigh) Anthropic | 90.5% ±1.8 | xhigh | 30 Jun 2026 | 1 Jul 2026 | Source ↗ | |
| 34 | Claude Opus 4.6 (32k thinking) Anthropic | 90.5% ±1.7 | 32K | 5 Feb 2026 | 6 Feb 2026 | Log ↗ | |
| 35 | GLM 5.3 Flash Open source Z.ai | 90.2% ±1.7 | max | 20 Aug 2026 | 26 Aug 2026 | Source ↗ | |
| 36 | Claude Opus 4.7 (xhigh) Anthropic | 90.2% ±1.8 | xhigh | 16 Apr 2026 | 17 Apr 2026 | Log ↗ | |
| 37 | GPT-5.6 Sol (low) OpenAI | 89.9% ±2.1 | low | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 38 | GLM 5.1 Open source Z.ai | 89.9% ±2.1 | — | 7 Apr 2026 | 10 Aug 2026 | Log ↗ | |
| 39 | GPT-5.4 OpenAI | 89.9% ±2.1 | high | 5 Mar 2026 | 15 Jul 2026 | Log ↗ | |
| 40 | Muse Spark Meta AI | 89.8% ±2.2 | — | 8 Apr 2026 | 8 Apr 2026 | Source ↗ | |
| 41 | Gemini 3 Flash Preview Google | 89.4% ±2.2 | high | 17 Dec 2025 | 6 Aug 2026 | Log ↗ | |
| 42 | Grok 4.20 SpaceXAI | 89.3% ±1.9 | — | 17 Feb 2026 | 13 Jul 2026 | Source ↗ | |
| 43 | GPT-5.4 (medium) OpenAI | 88.9% ±2.2 | medium | 5 Mar 2026 | 15 Jul 2026 | Log ↗ | |
| 44 | Grok 4.3 Beta xAI | 88.8% ±2.0 | high | 17 Apr 2026 | 17 Jun 2026 | Source ↗ | |
| 45 | Claude Opus 4.6 (64k thinking) Anthropic | 88.8% ±1.9 | 64K | 5 Feb 2026 | 6 Feb 2026 | Log ↗ | |
| 46 | Inkling Small (xhigh) Open source Thinking Machines | 88.5% ±1.9 | xhigh | 15 Jul 2026 | 14 Aug 2026 | Source ↗ | |
| 47 | Qwen 3.6 Plus (2026-04-02) Alibaba | 88.4% ±2.3 | — | 31 Mar 2026 | 7 Aug 2026 | Log ↗ | |
| 48 | Claude Opus 4.6 Anthropic | 88.4% ±2.3 | max | 5 Feb 2026 | 6 Aug 2026 | Log ↗ | |
| 49 | Inkling (xhigh) Open source Thinking Machines | 88.3% ±1.9 | xhigh | 15 Jul 2026 | 5 Aug 2026 | Source ↗ | |
| 50 | GPT-5.2 OpenAI | 88.2% ±1.9 | high | 11 Dec 2025 | 11 Dec 2025 | Log ↗ | |
| 51 | Claude Opus 5 (low) Anthropic | 87.9% ±2.3 | low | 24 Jul 2026 | 6 Aug 2026 | Log ↗ | |
| 52 | Kimi K2.7 Code Open source MoonshotAI | 87.9% ±2.3 | — | 12 Jun 2026 | 7 Aug 2026 | Log ↗ | |
| 53 | Qwen3.7 Plus Qwen | 87.9% ±2.3 | — | 2 Jun 2026 | 7 Aug 2026 | Log ↗ | |
| 54 | GPT-5.2 (medium) OpenAI | 87.9% ±1.9 | medium | 11 Dec 2025 | 11 Dec 2025 | Log ↗ | |
| 55 | GLM 5 Open source Z.ai | 87.8% ±2.3 | — | 11 Feb 2026 | 12 Feb 2026 | Log ↗ | |
| 56 | GPT-5.1 OpenAI | 87.6% ±1.9 | high | 13 Nov 2025 | 13 Nov 2025 | Log ↗ | |
| 57 | Kimi K2.5 (Fireworks) Open source Moonshot | 87.6% ±1.9 | — | 27 Jan 2026 | 2 Feb 2026 | Log ↗ | |
| 58 | GPT-5.6 Terra (low) OpenAI | 87.4% ±2.4 | low | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 59 | Qwen 3.6 Max (Preview) Alibaba | 87.4% ±2.4 | — | 20 Apr 2026 | 7 Aug 2026 | Log ↗ | |
| 60 | Claude Sonnet 4.6 (32k thinking) Anthropic | 87.4% ±2.0 | 32K | 17 Feb 2026 | 20 Feb 2026 | Log ↗ | |
| 61 | Grok 4 xAI | 87.0% ±2.0 | — | 9 Jul 2025 | — | Source ↗ | |
| 62 | GPT-5.4 mini (xhigh) OpenAI | 86.9% ±2.4 | xhigh | 17 Mar 2026 | 7 Aug 2026 | Log ↗ | |
| 63 | Claude Opus 4.7 Anthropic | 86.4% ±2.4 | max | 16 Apr 2026 | 6 Aug 2026 | Log ↗ | |
| 64 | Qwen3.5 397B A17B Open source Qwen | 86.4% ±2.4 | none | 13 Feb 2026 | 7 Aug 2026 | Log ↗ | |
| 65 | GPT-5 OpenAI | 86.2% ±2.1 | high | 7 Aug 2025 | 29 Oct 2025 | Log ↗ | |
| 66 | Claude Opus 4.5 (32k thinking) Anthropic | 86.0% ±2.1 | 32K | 24 Nov 2025 | 24 Nov 2025 | Log ↗ | |
| 67 | Claude Fable 5 Anthropic | 85.9% ±2.5 | max | 9 Jun 2026 | 6 Aug 2026 | Log ↗ | |
| 68 | Qwen3.6 27B Open source Qwen | 85.9% ±2.5 | — | 22 Apr 2026 | 7 Aug 2026 | Log ↗ | |
| 69 | Claude Opus 4.5 (16k thinking) Anthropic | 85.5% ±2.2 | 16K | 24 Nov 2025 | 25 Nov 2025 | Log ↗ | |
| 70 | Nemotron 3 Ultra Open source NVIDIA | 85.4% ±2.5 | — | 4 Jun 2026 | 10 Aug 2026 | Log ↗ | |
| 71 | GPT-5 (medium) OpenAI | 85.4% ±2.1 | medium | 7 Aug 2025 | 7 Aug 2025 | Log ↗ | |
| 72 | Gemini 2.5 Pro (Jun 2025) Google DeepMind | 85.3% ±2.1 | — | 5 Jun 2025 | 16 Nov 2025 | Log ↗ | |
| 73 | GPT-5.1 (medium) OpenAI | 85.0% ±2.1 | medium | 13 Nov 2025 | 17 Nov 2025 | Log ↗ | |
| 74 | Kimi K3 (low) Open source Moonshot | 84.8% ±2.6 | low | 16 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 75 | Qwen3.6 35B A3B Open source Qwen | 84.8% ±2.6 | none | 27 Apr 2026 | 7 Aug 2026 | Log ↗ | |
| 76 | GPT-5.4 (low) OpenAI | 84.8% ±2.6 | low | 5 Mar 2026 | 15 Jul 2026 | Log ↗ | |
| 77 | Qwen3.5 Plus 2026-02-15 Qwen | 84.8% ±2.6 | — | 16 Feb 2026 | 7 Aug 2026 | Log ↗ | |
| 78 | Gemini 2.5 Pro Preview (Jun 2025) Google DeepMind | 84.8% ±2.6 | — | 5 Jun 2025 | 5 Jun 2025 | Log ↗ | |
| 79 | Kimi K2 Thinking Turbo Open source Moonshot | 84.2% ±2.1 | — | 6 Nov 2025 | 11 Nov 2025 | Log ↗ | |
| 80 | Gemini 2.5 Pro Exp (Mar 2025) Google DeepMind | 83.8% ±2.6 | — | 25 Mar 2025 | 31 Mar 2025 | Log ↗ | |
| 81 | GPT-5.4 Mini OpenAI | 83.6% ±2.2 | high | 17 Mar 2026 | 15 Apr 2026 | Log ↗ | |
| 82 | Qwen3.5-35B-A3B Open source Qwen | 83.5% ±2.2 | — | 24 Feb 2026 | 28 Aug 2026 | Log ↗ | |
| 83 | DeepSeek V3.2 Open source DeepSeek | 83.4% ±2.0 | — | 1 Dec 2025 | 16 Dec 2025 | Log ↗ | |
| 84 | Gemini 3.5 Flash Lite Google | 83.3% ±2.7 | high | 21 Jul 2026 | 6 Aug 2026 | Log ↗ | |
| 85 | Qwen 3.6 Flash (2026-04-16) Alibaba | 83.3% ±2.7 | — | 27 Apr 2026 | 7 Aug 2026 | Log ↗ | |
| 86 | Claude Sonnet 4.6 Anthropic | 83.3% ±2.7 | high | 17 Feb 2026 | 13 Jul 2026 | Log ↗ | |
| 87 | Claude Sonnet 4.6 (medium) Anthropic | 83.3% ±2.7 | medium | 17 Feb 2026 | 13 Jul 2026 | Log ↗ | |
| 88 | GLM 4.7 Open source Z.ai | 83.3% ±2.4 | — | 22 Dec 2025 | 29 Jan 2026 | Log ↗ | |
| 89 | GPT-5.6 Sol (none) OpenAI | 82.8% ±2.7 | none | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 90 | GPT-5.2 (low) OpenAI | 82.7% ±2.3 | low | 11 Dec 2025 | 11 Dec 2025 | Log ↗ | |
| 91 | GPT-5.5 Instant OpenAI | 82.5% ±2.3 | — | 5 May 2026 | 2 Aug 2026 | Source ↗ | |
| 92 | Qwen3.7 Flash Qwen | 82.3% ±2.7 | — | 27 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 93 | GPT-5.6 Luna (low) OpenAI | 82.3% ±2.7 | low | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 94 | Qwen3.5-Flash Qwen | 82.3% ±2.7 | — | 25 Feb 2026 | 7 Aug 2026 | Log ↗ | |
| 95 | Claude Sonnet 4.5 (59k thinking) Anthropic | 82.3% ±2.7 | 59K | 29 Sep 2025 | 28 Oct 2025 | Log ↗ | |
| 96 | Gemini 3.1 Flash Lite Google | 81.8% ±2.7 | high | 3 Mar 2026 | 6 Aug 2026 | Log ↗ | |
| 97 | o3 OpenAI | 81.8% ±2.1 | high | 16 Apr 2025 | 16 Apr 2025 | Log ↗ | |
| 98 | Claude Sonnet 4.5 (32k thinking) Anthropic | 81.7% ±2.8 | 32K | 29 Sep 2025 | 21 Oct 2025 | Log ↗ | |
| 99 | o3 (medium) OpenAI | 80.8% ±2.8 | medium | 16 Apr 2025 | 7 Aug 2026 | Log ↗ | |
| 100 | Claude Opus 4.5 (no thinking) Anthropic | 80.7% ±2.4 | — | 24 Nov 2025 | 24 Nov 2025 | Log ↗ | |
| 101 | Claude Sonnet 5 Anthropic | 80.3% ±2.8 | max | 30 Jun 2026 | 6 Aug 2026 | Log ↗ | |
| 102 | Qwen3-235B-A22B-Thinking (Jul 2025) Open source Alibaba | 80.1% ±2.6 | — | 25 Jul 2025 | 11 Dec 2025 | Log ↗ | |
| 103 | o3 (low) OpenAI | 79.8% ±2.9 | low | 16 Apr 2025 | 15 Jul 2026 | Log ↗ | |
| 104 | Claude 3.7 Sonnet (64k thinking) Anthropic | 79.7% ±2.7 | 64K | 24 Feb 2025 | 26 May 2025 | Log ↗ | |
| 105 | o4 Mini OpenAI | 79.6% ±2.4 | high | 16 Apr 2025 | 16 Apr 2025 | Log ↗ | |
| 106 | Claude Sonnet 4 Anthropic | 79.2% ±2.7 | 59K | 22 May 2025 | 26 May 2025 | Log ↗ | |
| 107 | Qwen3.5-9B Open source Qwen | 79.0% ±2.4 | — | 24 Feb 2026 | 27 Aug 2026 | Log ↗ | |
| 108 | Claude Fable 5 (low) Anthropic | 78.8% ±2.9 | low | 9 Jun 2026 | 6 Aug 2026 | Log ↗ | |
| 109 | Claude Sonnet 4.5 (16k thinking) Anthropic | 78.8% ±2.9 | 16K | 29 Sep 2025 | 28 Oct 2025 | Log ↗ | |
| 110 | GPT-5.4 Nano OpenAI | 78.5% ±2.4 | high | 17 Mar 2026 | 14 Apr 2026 | Log ↗ | |
| 111 | o4-mini (medium) OpenAI | 77.8% ±3.0 | medium | 16 Apr 2025 | 7 Aug 2026 | Log ↗ | |
| 112 | GPT-5.6 Terra (none) OpenAI | 77.3% ±3.0 | none | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 113 | GPT-5.5 (no thinking) OpenAI | 77.3% ±3.0 | none | 23 Apr 2026 | 7 Aug 2026 | Log ↗ | |
| 114 | Claude Opus 4.1 Anthropic | 77.3% ±3.0 | 16K | 5 Aug 2025 | 5 Aug 2025 | Log ↗ | |
| 115 | o3 Mini OpenAI | 77.0% ±2.6 | high | 31 Jan 2025 | 13 Feb 2025 | Log ↗ | |
| 116 | Claude 3.7 Sonnet (32k thinking) Anthropic | 76.8% ±3.0 | 32K | 24 Feb 2025 | 10 Mar 2025 | Log ↗ | |
| 117 | Claude 3.7 Sonnet (16k thinking) Anthropic | 76.8% ±3.0 | 16K | 24 Feb 2025 | 26 Feb 2025 | Log ↗ | |
| 118 | o1 OpenAI | 76.8% ±3.0 | high | 17 Dec 2024 | 13 Feb 2025 | Log ↗ | |
| 119 | DeepSeek-R1 (May 2025) Open source DeepSeek | 76.3% ±2.4 | — | 28 May 2025 | 29 May 2025 | Log ↗ | |
| 120 | Grok-3 mini xAI | 76.3% ±3.0 | low | 24 Jun 2025 | 10 Apr 2025 | Log ↗ | |
| 121 | Claude Opus 4 Anthropic | 76.3% ±3.0 | 16K | 22 May 2025 | 22 May 2025 | Log ↗ | |
| 122 | Gemma 4 31B Open source Google | 75.8% ±3.1 | minimal | 2 Apr 2026 | 6 Aug 2026 | Log ↗ | |
| 123 | gpt-oss-120b Open source OpenAI | 75.8% ±2.7 | high | 5 Aug 2025 | 11 Dec 2025 | Log ↗ | |
| 124 | Grok 3 xAI | 75.8% ±2.7 | — | 9 Apr 2025 | 26 May 2025 | Log ↗ | |
| 125 | o1 (medium) OpenAI | 75.8% ±3.1 | medium | 17 Dec 2024 | 27 Jan 2025 | Log ↗ | |
| 126 | o4-mini (low) OpenAI | 75.3% ±3.1 | low | 16 Apr 2025 | 11 Jul 2026 | Log ↗ | |
| 127 | GPT-5 Mini OpenAI | 75.0% ±2.3 | high | 7 Aug 2025 | 30 Oct 2025 | Log ↗ | |
| 128 | GPT-5.4 (none) OpenAI | 74.7% ±3.1 | none | 5 Mar 2026 | 15 Jul 2026 | Log ↗ | |
| 129 | o3-mini (medium) OpenAI | 74.3% ±2.5 | medium | 31 Jan 2025 | 31 Jan 2025 | Log ↗ | |
| 130 | o1 (low) OpenAI | 74.2% ±3.1 | low | 17 Dec 2024 | 20 Jul 2026 | Log ↗ | |
| 131 | Claude Sonnet 4.5 (no thinking) Anthropic | 73.7% ±3.1 | — | 29 Sep 2025 | 29 Sep 2025 | Log ↗ | |
| 132 | DeepSeek v4 Pro (unknown thinking) Open source DeepSeek | 73.2% ±3.2 | none | 24 Apr 2026 | 6 Aug 2026 | Log ↗ | |
| 133 | Gemma 4 26B A4B Open source Google | 73.2% ±3.2 | minimal | 2 Apr 2026 | 6 Aug 2026 | Log ↗ | |
| 134 | GPT-5.2 (none) OpenAI | 73.2% ±3.2 | none | 11 Dec 2025 | 13 Jul 2026 | Log ↗ | |
| 135 | Qwen3-Max-Instruct Alibaba | 72.6% ±2.7 | — | 24 Sep 2025 | 6 Oct 2025 | Log ↗ | |
| 136 | GPT-5.4 nano (low) OpenAI | 72.2% ±3.2 | low | 17 Mar 2026 | 7 Aug 2026 | Log ↗ | |
| 137 | GPT-5 mini (minimal) OpenAI | 71.7% ±3.2 | minimal | 7 Aug 2025 | 7 Aug 2026 | Log ↗ | |
| 138 | GPT-5 (minimal) OpenAI | 71.7% ±3.2 | minimal | 7 Aug 2025 | 20 Jul 2026 | Log ↗ | |
| 139 | R1 Open source DeepSeek | 71.7% ±3.1 | — | 20 Jan 2025 | 26 May 2025 | Log ↗ | |
| 140 | GPT-5 mini (medium) OpenAI | 71.7% ±2.3 | medium | 7 Aug 2025 | 7 Aug 2025 | Log ↗ | |
| 141 | seed-oss-36b-instruct | 71.5% ±2.7 | — | 20 Aug 2025 | 30 Aug 2026 | Log ↗ | |
| 142 | Claude Haiku 4.5 Anthropic | 71.2% ±3.2 | 32K | 15 Oct 2025 | 22 Oct 2025 | Log ↗ | |
| 143 | Qwen3 235B A22B Open source Qwen | 70.7% ±2.7 | — | 28 Apr 2025 | 3 Jun 2025 | Log ↗ | |
| 144 | Qwen3-30B-A3B-Thinking (Jul 2025) Open source Alibaba | 70.1% ±2.8 | — | 30 Jul 2025 | 30 Aug 2026 | Log ↗ | |
| 145 | GPT-5 Nano OpenAI | 69.4% ±2.8 | high | 7 Aug 2025 | 30 Oct 2025 | Log ↗ | |
| 146 | GPT-4.5 Preview (Feb 2025) OpenAI | 68.7% ±3.3 | — | 27 Feb 2025 | 28 Feb 2025 | Log ↗ | |
| 147 | DeepSeek-V3 (Mar 2025) Open source DeepSeek | 67.6% ±2.7 | — | 24 Mar 2025 | 1 Apr 2025 | Log ↗ | |
| 148 | GPT-5 nano (medium) OpenAI | 67.4% ±2.7 | medium | 7 Aug 2025 | 7 Aug 2025 | Log ↗ | |
| 149 | Llama 4 Maverick (FP8) Open source Meta AI | 67.0% ±2.8 | — | 5 Apr 2025 | 8 Apr 2025 | Log ↗ | |
| 150 | GPT-4.1 OpenAI | 66.9% ±2.8 | — | 14 Apr 2025 | 14 Apr 2025 | Log ↗ | |
| 151 | GPT-5.1 (no thinking) OpenAI | 66.7% ±3.4 | none | 13 Nov 2025 | 7 Aug 2026 | Log ↗ | |
| 152 | Claude 3.7 Sonnet Anthropic | 66.0% ±2.7 | — | 24 Feb 2025 | 24 Feb 2025 | Log ↗ | |
| 153 | GPT-4.1 Mini OpenAI | 65.8% ±2.7 | — | 14 Apr 2025 | 14 Apr 2025 | Log ↗ | |
| 154 | Qwen3 32B Open source Qwen | 65.7% ±2.8 | — | 29 Apr 2025 | 28 Aug 2026 | Log ↗ | |
| 155 | Gemini 2.0 Pro Exp (Feb 2025) Google DeepMind | 65.7% ±3.4 | — | 5 Feb 2025 | 7 Feb 2025 | Log ↗ | |
| 156 | QWQ-Plus Alibaba | 65.4% ±2.8 | — | 8 Apr 2025 | 11 Apr 2025 | Log ↗ | |
| 157 | QwQ-32B Open source Alibaba | 65.3% ±2.8 | — | 5 Mar 2025 | 28 Aug 2026 | Log ↗ | |
| 158 | GPT-5.4 mini (none) OpenAI | 64.1% ±3.4 | none | 17 Mar 2026 | 7 Aug 2026 | Log ↗ | |
| 159 | Gemini 2.0 Flash (Feb 2025) Google DeepMind,Google | 64.1% ±2.7 | — | 5 Feb 2025 | 6 Feb 2025 | Log ↗ | |
| 160 | DeepSeek-R1-Distill-Qwen-32B Open source DeepSeek | 64.1% ±2.8 | — | 20 Jan 2025 | 28 Aug 2026 | Log ↗ | |
| 161 | Qwen3 14B Open source Qwen | 63.8% ±2.9 | — | 29 Apr 2025 | 28 Aug 2026 | Log ↗ | |
| 162 | GPT-5.6 Luna (none) OpenAI | 63.6% ±3.4 | none | 9 Jul 2026 | 7 Aug 2026 | Log ↗ | |
| 163 | o1-mini OpenAI | 62.4% ±2.7 | high | 12 Sep 2024 | 13 Feb 2025 | Log ↗ | |
| 164 | Qwen3 30B A3B Open source Qwen | 61.7% ±3.0 | — | 29 Apr 2025 | 30 Aug 2026 | Log ↗ | |
| 165 | gpt-oss-20b Open source OpenAI | 60.8% ±2.5 | medium | 5 Aug 2025 | 27 Aug 2026 | Log ↗ | |
| 166 | GLM 4.7 Flash Open source Z.ai | 60.5% ±2.9 | — | 19 Jan 2026 | 30 Aug 2026 | Log ↗ | |
| 167 | Mistral Medium 3 Mistral | 59.5% ±2.8 | — | 7 May 2025 | 7 May 2025 | Log ↗ | |
| 168 | o1-mini (medium) OpenAI | 59.5% ±2.8 | medium | 12 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 169 | GPT-5 nano (low) OpenAI | 57.6% ±3.5 | low | 7 Aug 2025 | 13 Jul 2026 | Log ↗ | |
| 170 | Gemini 1.5 Pro (Sept 2024) Google DeepMind | 57.2% ±2.8 | — | 24 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 171 | Gemini 2.0 Flash Thinking Exp Google DeepMind,Google | 57.1% ±3.5 | — | 21 Jan 2025 | 6 Feb 2025 | Log ↗ | |
| 172 | Qwen3 8B Open source Qwen | 56.8% ±2.9 | — | 28 Apr 2025 | 27 Aug 2026 | Log ↗ | |
| 173 | DeepSeek V3 Open source DeepSeek | 56.5% ±2.8 | — | 26 Dec 2024 | 27 Jan 2025 | Log ↗ | |
| 174 | Qwen2.5-Max Alibaba | 56.1% ±2.8 | — | 25 Jan 2025 | 1 Apr 2025 | Log ↗ | |
| 175 | Magistral Small 1.0 Open source Mistral AI | 56.1% ±3.5 | — | 10 Jun 2025 | 6 Aug 2026 | Log ↗ | |
| 176 | Phi 4 Open source Microsoft | 56.1% ±2.6 | — | 12 Dec 2024 | 31 Jan 2025 | Log ↗ | |
| 177 | R1 Distill Llama 70B Open source DeepSeek | 55.7% ±3.0 | — | 23 Jan 2025 | 10 Mar 2025 | Log ↗ | |
| 178 | Qwen3-30B-A3B-Instruct (Jul 2025) Open source Alibaba | 55.6% ±2.9 | — | 29 Jul 2025 | 30 Aug 2026 | Log ↗ | |
| 179 | GPT-5.4 nano (no thinking) OpenAI | 55.6% ±3.5 | none | 17 Mar 2026 | 7 Aug 2026 | Log ↗ | |
| 180 | Claude 3.5 Sonnet (Oct 2024) Anthropic | 55.3% ±2.8 | — | 22 Oct 2024 | 27 Jan 2025 | Log ↗ | |
| 181 | Claude 3.5 Sonnet (Jun 2024) Anthropic | 54.0% ±2.8 | — | 20 Jun 2024 | 27 Jan 2025 | Log ↗ | |
| 182 | Grok-2 (Dec 2024) xAI | 53.8% ±2.7 | — | 12 Dec 2024 | 27 Jan 2025 | Log ↗ | |
| 183 | Qwen3-4B Open source Alibaba | 52.3% ±3.0 | — | 29 Apr 2025 | 28 Aug 2026 | Log ↗ | |
| 184 | Llama 4 Scout Open source Meta | 51.8% ±3.2 | — | 5 Apr 2025 | 8 Apr 2025 | Log ↗ | |
| 185 | Mistral Large 2 (Nov 2024) Open source Mistral AI | 51.3% ±2.7 | — | 18 Nov 2024 | 25 Feb 2025 | Log ↗ | |
| 186 | Llama 3.1-405B Open source Meta AI | 50.9% ±2.6 | — | 23 Jul 2024 | 27 Jan 2025 | Log ↗ | |
| 187 | o1-preview OpenAI | 50.3% ±2.8 | — | 12 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 188 | GPT-4o (Aug 2024) OpenAI | 49.2% ±2.6 | — | 6 Aug 2024 | 27 Jan 2025 | Log ↗ | |
| 189 | Qwen2.5-72B Open source Alibaba | 49.1% ±2.7 | — | 19 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 190 | Mistral Small 3.2 Open source Mistral AI | 49.1% ±2.8 | — | 20 Jun 2025 | 28 Aug 2026 | Log ↗ | |
| 191 | Mistral Large 2 (Jul 2024) Open source Mistral AI | 49.0% ±2.6 | — | 24 Jul 2024 | 27 Jan 2025 | Log ↗ | |
| 192 | GPT-4.1 Nano OpenAI | 48.9% ±2.5 | — | 14 Apr 2025 | 14 Apr 2025 | Log ↗ | |
| 193 | GPT-4o (May 2024) OpenAI | 48.9% ±2.6 | — | 13 May 2024 | 27 Jan 2025 | Log ↗ | |
| 194 | GPT-5 nano (minimal) OpenAI | 48.5% ±3.6 | minimal | 7 Aug 2025 | 13 Jul 2026 | Log ↗ | |
| 195 | Qwen Plus (Jan 2025) Alibaba | 48.1% ±2.7 | — | 25 Jan 2025 | 7 Apr 2025 | Log ↗ | |
| 196 | GPT-4o (Nov 2024) OpenAI | 47.9% ±2.6 | — | 20 Nov 2024 | 5 Feb 2025 | Log ↗ | |
| 197 | Gemma 3 27B Open source Google | 47.7% ±2.7 | — | 12 Mar 2025 | 27 Aug 2026 | Log ↗ | |
| 198 | Magistral Small 1.2 Open source Mistral AI | 47.6% ±2.6 | — | 18 Sep 2025 | 30 Aug 2026 | Log ↗ | |
| 199 | Mistral Small 3.1 Open source Mistral AI | 47.5% ±2.8 | — | 17 Mar 2025 | 18 Mar 2025 | Log ↗ | |
| 200 | Llama 3.3 70B Open source Meta AI | 47.4% ±2.9 | — | 6 Dec 2024 | 27 Jan 2025 | Log ↗ | |
| 201 | Gemini 1.5 Flash (Sep 2024) Google DeepMind | 47.3% ±2.8 | — | 24 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 202 | Mistral Small 3 Open source Mistral | 47.3% ±2.8 | — | 30 Jan 2025 | 30 Aug 2026 | Log ↗ | |
| 203 | Claude 3 Opus Anthropic | 47.2% ±2.6 | — | 29 Feb 2024 | 27 Jan 2025 | Log ↗ | |
| 204 | GPT-4 Turbo (Apr 2024) OpenAI | 46.6% ±2.7 | — | 9 Apr 2024 | 27 Feb 2025 | Log ↗ | |
| 205 | Tulu 3 (Tülu 3) 70B Open source Allen Institute for AI,University of Washington | 46.3% ±2.4 | — | 21 Nov 2024 | 27 Jan 2025 | Log ↗ | |
| 206 | Qwen2.5-32B Open source Alibaba | 46.1% ±2.6 | — | 17 Sep 2024 | 30 Jan 2025 | Log ↗ | |
| 207 | Gemini 1.5 Pro (May 2024) Google DeepMind | 45.9% ±2.6 | — | 14 May 2024 | 27 Jan 2025 | Log ↗ | |
| 208 | qwen3-4b-instruct-2507 | 45.8% ±2.9 | — | 6 Aug 2025 | 28 Aug 2026 | Log ↗ | |
| 209 | DeepSeek-R1-Distill-Qwen-14B Open source DeepSeek | 44.7% ±2.9 | — | 20 Jan 2025 | 10 Mar 2025 | Log ↗ | |
| 210 | Llama 3.1-70B Open source Meta AI | 44.2% ±2.5 | — | 23 Jul 2024 | 27 Jan 2025 | Log ↗ | |
| 211 | WizardLM-2 8x22B Open source microsoft | 43.4% ±2.3 | — | 16 Apr 2024 | 27 Jan 2025 | Log ↗ | |
| 212 | GPT-4 Turbo Preview (Nov 2023) OpenAI | 42.4% ±2.4 | — | 6 Nov 2023 | 27 Jan 2025 | Log ↗ | |
| 213 | GPT-4 Turbo Preview (January 2024) OpenAI | 42.3% ±2.7 | — | 25 Jan 2024 | 27 Jan 2025 | Log ↗ | |
| 214 | Qwen-Turbo Alibaba | 41.8% ±2.6 | — | 1 Nov 2024 | 7 Apr 2025 | Log ↗ | |
| 215 | Llama 3.2 90B Open source Meta AI | 41.0% ±2.3 | — | 24 Sep 2024 | 27 Jan 2025 | Log ↗ | |
| 216 | Qwen2-72B Open source Alibaba | 40.8% ±2.6 | — | 7 Jun 2024 | 27 Jan 2025 | Log ↗ | |
| 217 | Claude 3 Sonnet Anthropic | 40.6% ±2.4 | — | 29 Feb 2024 | 27 Jan 2025 | Log ↗ | |
| 218 | Llama 3-70B Open source Meta AI | 40.6% ±2.6 | — | 18 Apr 2024 | 27 Jan 2025 | Log ↗ | |
| 219 | Gemini 1.5 Flash (May 2024) Google DeepMind | 40.4% ±2.4 | — | 23 May 2024 | 27 Jan 2025 | Log ↗ | |
| 220 | Gemma 3 12B Open source Google | 39.5% ±2.5 | — | 12 Mar 2025 | 28 Aug 2026 | Log ↗ | |
| 221 | Mistral Large mistralai | 38.8% ±2.5 | — | 26 Feb 2024 | 27 Jan 2025 | Log ↗ | |
| 222 | Claude 3.5 Haiku (Oct 2024) Anthropic | 38.1% ±2.6 | — | 22 Oct 2024 | 12 Mar 2025 | Log ↗ | |
| 223 | Qwen3-1.7B Open source Alibaba | 38.0% ±2.7 | — | 29 Apr 2025 | 28 Aug 2026 | Log ↗ | |
| 224 | GPT-4o-mini OpenAI | 37.7% ±2.4 | — | 18 Jul 2024 | 27 Jan 2025 | Log ↗ | |
| 225 | Hermes 2 Theta Llama-3 70B Open source Nous Research,Arcee AI | 37.5% ±2.5 | — | 20 Jun 2024 | 27 Jan 2025 | Log ↗ | |
| 226 | Gemma 2 27B Open source Google | 36.5% ±2.4 | — | 24 Jun 2024 | 27 Jan 2025 | Log ↗ | |
| 227 | Claude 3 Haiku Anthropic | 36.3% ±2.4 | — | 7 Mar 2024 | 27 Jan 2025 | Log ↗ | |
| 228 | GPT-4 (Mar 2023) OpenAI | 35.7% ±2.4 | — | 14 Mar 2023 | 23 Oct 2025 | Log ↗ | |
| 229 | Qwen2.5-7B Open source Alibaba | 35.5% ±2.3 | — | 19 Sep 2024 | 30 Aug 2026 | Log ↗ | |
| 230 | Claude 2 Anthropic | 34.7% ±2.5 | — | 11 Jul 2023 | 27 Jan 2025 | Log ↗ | |
| 231 | Mixtral 8x22B Open source Mistral AI | 34.1% ±2.1 | — | 17 Apr 2024 | 27 Jan 2025 | Log ↗ | |
| 232 | Gemini 1.0 Pro Google DeepMind | 34.0% ±2.0 | — | 13 Dec 2023 | 27 Jan 2025 | Log ↗ | |
| 233 | Eurus-2-7B-PRIME Open source Tsinghua University,University of Illinois Urbana-Champaign (UIUC),Shanghai AI Lab,Peking University,Shanghai Jiao Tong University,CUHK Shenzhen Research Institute | 33.9% ±2.1 | — | 31 Dec 2024 | 27 Jan 2025 | Log ↗ | |
| 234 | DeepSeek-R1-Distill-Qwen-1.5B Open source DeepSeek | 33.6% ±2.1 | — | 20 Jan 2025 | 28 Aug 2026 | Log ↗ | |
| 235 | Gemini 1.5 Flash 8B Google DeepMind | 33.0% ±2.4 | — | 3 Oct 2024 | 12 Mar 2025 | Log ↗ | |
| 236 | Claude 2.1 Anthropic | 33.0% ±2.3 | — | 21 Nov 2023 | 27 Jan 2025 | Log ↗ | |
| 237 | DBRX Open source Databricks | 32.9% ±3.3 | — | 27 Mar 2024 | 27 Jan 2025 | Log ↗ | |
| 238 | Yi-1.5-34B Open source 01.AI | 32.0% ±2.0 | — | 13 May 2024 | 27 Jan 2025 | Log ↗ | |
| 239 | Qwen1.5-32B Open source Alibaba | 30.7% ±2.0 | — | 3 Apr 2024 | 27 Jan 2025 | Log ↗ | |
| 240 | GPT-4 (Jun 2023) OpenAI | 30.7% ±2.3 | — | 13 Jun 2023 | 27 Jan 2025 | Log ↗ | |
| 241 | Mixtral 8x7B Open source Mistral AI | 30.6% ±1.9 | — | 11 Dec 2023 | 27 Jan 2025 | Log ↗ | |
| 242 | Mistral Nemo Open source Mistral | 29.9% ±2.1 | — | 18 Jul 2024 | 27 Jan 2025 | Log ↗ | |
| 243 | Qwen1.5-72B Open source Alibaba | 28.8% ±2.0 | — | 4 Feb 2024 | 27 Jan 2025 | Log ↗ | |
| 244 | Granite 4.0 Micro Open source IBM | 28.3% ±1.8 | — | 20 Oct 2025 | 30 Aug 2026 | Log ↗ | |
| 245 | GPT-3.5 Turbo (Nov 2023) OpenAI | 28.0% ±1.9 | — | 6 Nov 2023 | 27 Jan 2025 | Log ↗ | |
| 246 | phi-3-medium 14B Open source Microsoft | 27.6% ±1.6 | — | 23 Apr 2024 | 31 Jan 2025 | Log ↗ | |
| 247 | Gemma 2 9B Open source Google DeepMind | 27.5% ±2.3 | — | 24 Jun 2024 | 27 Jan 2025 | Log ↗ | |
| 248 | GPT-3.5 Turbo (Jan 2024) OpenAI | 27.2% ±2.0 | — | 25 Jan 2024 | 27 Jan 2025 | Log ↗ | |
| 249 | Ministral 8B Open source Mistral AI | 27.1% ±2.1 | — | 16 Oct 2024 | 27 Jan 2025 | Log ↗ | |
| 250 | Llama 3.1-8B Open source Meta AI | 27.0% ±1.9 | — | 23 Jul 2024 | 27 Aug 2026 | Log ↗ | |
| 251 | Llama 2-70B Open source Meta AI | 26.3% ±2.0 | — | 18 Jul 2023 | 27 Jan 2025 | Log ↗ | |
| 252 | Llama 3-8B Open source Meta AI | 26.1% ±1.7 | — | 18 Apr 2024 | 27 Jan 2025 | Log ↗ | |
| 253 | Ministral 3B Mistral AI | 25.3% ±1.9 | — | 16 Oct 2024 | 27 Jan 2025 | Log ↗ | |
| 254 | DeepSeek LLM 67B Open source DeepSeek | 24.6% ±1.4 | — | 29 Nov 2023 | 27 Jan 2025 | Log ↗ | |
| 255 | granite-4.0-1b | 24.0% ±1.6 | — | 28 Oct 2025 | 28 Aug 2026 | Log ↗ | |
| 256 | Llama 3.2 1B Open source Meta AI | 23.9% ±1.6 | — | 24 Sep 2024 | 30 Aug 2026 | Log ↗ | |
| 257 | Gemma 3 4B Open source Google | 23.2% ±1.9 | — | 12 Mar 2025 | 28 Aug 2026 | Log ↗ | |
| 258 | Gemma 3 1B Open source Google DeepMind | 19.9% ±1.5 | — | 12 Mar 2025 | 28 Aug 2026 | Log ↗ | |
| 259 | Mistral 7B v0.3 Open source Mistral AI | 15.2% ±1.0 | — | 27 May 2024 | 27 Jan 2025 | Log ↗ | |
| 260 | Yi-34B Open source 01.AI | 14.7% ±1.1 | — | 2 Nov 2023 | 27 Jan 2025 | Log ↗ | |
| 261 | granite-4.0-350m | 11.2% ±0.9 | — | 28 Oct 2025 | 28 Aug 2026 | Log ↗ | |
| 262 | deepseek-r1-0528-qwen3-8b | 9.3% ±1.1 | — | 28 May 2025 | 28 Aug 2026 | Log ↗ |
Caveats: settings (reasoning effort, agent scaffold) differ between rows and materially affect scores; the best reported setting per model is shown. Source: Epoch AI, CC BY 4.0.