| # | Model | Score | Relative | Setting | Released | Run | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | GPT-6 Astra OpenAI | 75.6% ±1.4 | max | 3 Sep 2026 | 30 Aug 2026 | Source ↗ | |
| 2 | Gemini 3.1 Pro Preview Google | 73.5% ±1.4 | high | 19 Feb 2026 | 10 Aug 2026 | Source ↗ | |
| 3 | Claude Fable 5.1 Anthropic | 70.8% ±1.4 | max | 1 Sep 2026 | 1 Sep 2026 | Source ↗ | |
| 4 | Claude Fable 5 (xhigh) Anthropic | 70.7% ±1.4 | xhigh | 9 Jun 2026 | 10 Aug 2026 | Source ↗ | |
| 5 | Gemini 3.8 Flash Google | 69.7% ±1.5 | high | 2 Sep 2026 | 2 Sep 2026 | Source ↗ | |
| 6 | GPT-5.6 Sol OpenAI | 69.7% ±1.5 | max | 9 Jul 2026 | 10 Aug 2026 | Source ↗ | |
| 7 | Gemini 3.7 Flash Google | 69.2% ±1.5 | high | 13 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 8 | Gemini 3 Flash Preview Google | 66.8% ±1.5 | high | 17 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 9 | Gemini 3.6 Flash Google | 66.2% ±1.5 | high | 21 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 10 | Gemini 3.5 Flash Google | 66.2% ±1.5 | high | 19 May 2026 | 27 Aug 2026 | Log ↗ | |
| 11 | GPT-5.5 (xhigh) OpenAI | 63.0% ±1.5 | xhigh | 23 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 12 | Muse Spark 1.2 (xhigh) Meta AI | 60.3% ±1.6 | xhigh | 5 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 13 | Claude Opus 5 Anthropic | 59.9% ±1.6 | max | 24 Jul 2026 | 10 Aug 2026 | Source ↗ | |
| 14 | Muse Spark 1.1 Meta | 57.8% ±1.6 | — | 9 Jul 2026 | 31 Aug 2026 | Log ↗ | |
| 15 | Qwen3.7 Max Qwen | 55.8% ±1.6 | — | 19 May 2026 | 27 Aug 2026 | Log ↗ | |
| 16 | Claude Opus 4.8 Anthropic | 53.0% ±1.6 | max | 28 May 2026 | 27 Aug 2026 | Log ↗ | |
| 17 | DeepSeek V4 Pro 0813 Open source DeepSeek | 52.9% ±1.6 | max | 13 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 18 | Qwen 3.6 Max (Preview) Alibaba | 52.0% ±1.6 | — | 20 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 19 | Claude Opus 4.7 (xhigh) Anthropic | 51.7% ±1.6 | xhigh | 16 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 20 | Kimi K3 Open source MoonshotAI | 50.6% ±1.6 | max | 16 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 21 | GPT-5 OpenAI | 50.1% ±1.6 | high | 7 Aug 2025 | 27 Aug 2026 | Log ↗ | |
| 22 | o3 OpenAI | 49.4% ±1.6 | high | 16 Apr 2025 | 27 Aug 2026 | Log ↗ | |
| 23 | Grok 4.6 SpaceXAI | 49.3% ±1.6 | high | 12 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 24 | Grok 4.6 (xhigh) xAI | 48.9% ±1.6 | xhigh | 12 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 25 | Qwen3-Max-Instruct Alibaba | 48.7% ±1.6 | — | 24 Sep 2025 | 27 Aug 2026 | Log ↗ | |
| 26 | Grok 4.5 SpaceXAI | 48.3% ±1.6 | high | 8 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 27 | GPT-5.1 OpenAI | 48.0% ±1.6 | high | 13 Nov 2025 | 27 Aug 2026 | Log ↗ | |
| 28 | Qwen3.8 Max (0902) (xhigh) Alibaba | 47.3% ±1.6 | xhigh | 1 Sep 2026 | 2 Sep 2026 | Source ↗ | |
| 29 | Claude Opus 4.6 Anthropic | 47.0% ±1.6 | max | 5 Feb 2026 | 27 Aug 2026 | Log ↗ | |
| 30 | DeepSeek v4 Pro (high) Open source DeepSeek | 47.0% ±1.6 | max | 24 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 31 | GPT-5.4 Pro (xhigh) OpenAI | 46.3% ±1.6 | xhigh | 5 Mar 2026 | 27 Aug 2026 | Log ↗ | |
| 32 | Qwen3.8 Max (xhigh) Alibaba | 45.8% ±1.6 | xhigh | 2 Aug 2026 | 27 Aug 2026 | Log ↗ | |
| 33 | Claude Opus 4.5 (32k thinking) Anthropic | 45.7% ±1.6 | 32K | 24 Nov 2025 | 27 Aug 2026 | Log ↗ | |
| 34 | GPT-5.4 (xhigh) OpenAI | 45.1% ±1.6 | xhigh | 5 Mar 2026 | 27 Aug 2026 | Log ↗ | |
| 35 | Qwen 3.6 Plus (2026-04-02) Alibaba | 44.1% ±1.6 | — | 31 Mar 2026 | 27 Aug 2026 | Log ↗ | |
| 36 | GPT-5.6 Terra OpenAI | 43.2% ±1.6 | max | 9 Jul 2026 | 10 Aug 2026 | Source ↗ | |
| 37 | o1 OpenAI | 41.1% ±1.6 | high | 17 Dec 2024 | 31 Aug 2026 | Log ↗ | |
| 38 | GLM 5.3 Open source Z.ai | 41.0% ±1.6 | max | 14 Aug 2026 | 28 Aug 2026 | Log ↗ | |
| 39 | GPT-5.6 Luna OpenAI | 41.0% ±1.6 | max | 9 Jul 2026 | 10 Aug 2026 | Source ↗ | |
| 40 | Qwen3-235B-A22B-Thinking (Jul 2025) Open source Alibaba | 40.4% ±1.6 | — | 25 Jul 2025 | 27 Aug 2026 | Log ↗ | |
| 41 | Inkling (xhigh) Open source Thinking Machines | 40.3% ±1.6 | xhigh | 15 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 42 | GPT-5.2 (xhigh) OpenAI | 37.1% ±1.5 | xhigh | 11 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 43 | Kimi K2.7 Code Open source MoonshotAI | 36.5% ±1.5 | — | 12 Jun 2026 | 27 Aug 2026 | Log ↗ | |
| 44 | Claude Sonnet 4.6 Anthropic | 35.5% ±1.5 | high | 17 Feb 2026 | 10 Aug 2026 | Source ↗ | |
| 45 | Kimi K2.6 Open source MoonshotAI | 34.9% ±1.5 | — | 20 Apr 2026 | 10 Aug 2026 | Source ↗ | |
| 46 | Kimi K2.5 Open source MoonshotAI | 34.3% ±1.5 | — | 27 Jan 2026 | 27 Aug 2026 | Log ↗ | |
| 47 | GPT-5.2 OpenAI | 34.3% ±1.5 | high | 11 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 48 | GLM 5.2 Open source Z.ai | 34.2% ±1.5 | max | 16 Jun 2026 | 27 Aug 2026 | Log ↗ | |
| 49 | GLM 5.1 Open source Z.ai | 34.0% ±1.5 | — | 7 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 50 | Claude Sonnet 5 Anthropic | 33.7% ±1.5 | max | 30 Jun 2026 | 27 Aug 2026 | Log ↗ | |
| 51 | DeepSeek V4 Flash 0731 Open source DeepSeek | 33.6% ±1.5 | max | 31 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 52 | Grok 4.3 Beta xAI | 33.2% ±1.5 | high | 17 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 53 | Claude Sonnet 5 (xhigh) Anthropic | 32.9% ±1.5 | xhigh | 30 Jun 2026 | 27 Aug 2026 | Log ↗ | |
| 54 | GPT-5.2 (low) OpenAI | 32.8% ±1.5 | low | 11 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 55 | GPT-5.2 (medium) OpenAI | 32.7% ±1.5 | medium | 11 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 56 | GLM 4.7 Open source Z.ai | 32.2% ±1.5 | — | 22 Dec 2025 | 27 Aug 2026 | Log ↗ | |
| 57 | GPT-4.1 OpenAI | 31.1% ±1.5 | — | 14 Apr 2025 | 31 Aug 2026 | Log ↗ | |
| 58 | Claude Sonnet 4.5 (59k thinking) Anthropic | 30.7% ±1.5 | 59K | 29 Sep 2025 | 27 Aug 2026 | Log ↗ | |
| 59 | Grok 4.20 SpaceXAI | 30.2% ±1.5 | — | 17 Feb 2026 | 27 Aug 2026 | Log ↗ | |
| 60 | GPT-5.4 Mini OpenAI | 29.4% ±1.4 | high | 17 Mar 2026 | 27 Aug 2026 | Log ↗ | |
| 61 | GPT-4o (Aug 2024) OpenAI | 26.0% ±1.4 | — | 6 Aug 2024 | 31 Aug 2026 | Log ↗ | |
| 62 | Qwen3.5 Plus 2026-02-15 Qwen | 25.4% ±1.4 | — | 16 Feb 2026 | 27 Aug 2026 | Log ↗ | |
| 63 | Claude Sonnet 4.5 (no thinking) Anthropic | 23.7% ±1.3 | — | 29 Sep 2025 | 10 Aug 2026 | Source ↗ | |
| 64 | GPT-5 Mini OpenAI | 21.6% ±1.3 | high | 7 Aug 2025 | 10 Aug 2026 | Source ↗ | |
| 65 | Qwen3.5-Flash Qwen | 20.3% ±1.3 | — | 25 Feb 2026 | 27 Aug 2026 | Log ↗ | |
| 66 | o4-mini (low) OpenAI | 19.6% ±1.3 | low | 16 Apr 2025 | 27 Aug 2026 | Log ↗ | |
| 67 | Inkling Small (xhigh) Open source Thinking Machines | 19.1% ±1.2 | xhigh | 15 Jul 2026 | 27 Aug 2026 | Log ↗ | |
| 68 | o4 Mini OpenAI | 18.8% ±1.2 | high | 16 Apr 2025 | 27 Aug 2026 | Log ↗ | |
| 69 | Qwen 3.6 Flash (2026-04-16) Alibaba | 15.9% ±1.2 | — | 27 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 70 | o3 Mini OpenAI | 15.3% ±1.1 | high | 31 Jan 2025 | 31 Aug 2026 | Log ↗ | |
| 71 | Claude Haiku 4.5 Anthropic | 13.2% ±1.1 | — | 15 Oct 2025 | 10 Aug 2026 | Source ↗ | |
| 72 | GPT-4.1 Mini OpenAI | 12.7% ±1.1 | — | 14 Apr 2025 | 31 Aug 2026 | Log ↗ | |
| 73 | Claude 3 Opus Anthropic | 12.6% ±1.0 | — | 29 Feb 2024 | 31 Aug 2026 | Log ↗ | |
| 74 | GPT-5.4 Nano OpenAI | 11.7% ±1.0 | high | 17 Mar 2026 | 27 Aug 2026 | Log ↗ | |
| 75 | GPT-5 Nano OpenAI | 11.7% ±1.0 | high | 7 Aug 2025 | 10 Aug 2026 | Source ↗ | |
| 76 | Gemma 4 31B Open source Google | 10.4% ±1.0 | — | 2 Apr 2026 | 27 Aug 2026 | Log ↗ | |
| 77 | GPT-4o-mini OpenAI | 8.3% ±0.9 | — | 18 Jul 2024 | 31 Aug 2026 | Log ↗ | |
| 78 | GPT-4.1 Nano OpenAI | 6.0% ±0.8 | — | 14 Apr 2025 | 31 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.