| # | Model | Score | Relative | Setting | Released | Run | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | GPT-6 Astra (xhigh) OpenAI | 74.1% | xhigh | 3 Sep 2026 | — | Source ↗ | |
| 2 | Gemini 3.8 Flash Google | 73.8% | high | 2 Sep 2026 | — | Source ↗ | |
| 3 | Claude Opus 5 Anthropic | 73.6% | max | 24 Jul 2026 | — | Source ↗ | |
| 4 | GPT-6 Astra OpenAI | 73.2% | high | 3 Sep 2026 | — | Source ↗ | |
| 5 | Claude Opus 5 (xhigh) Anthropic | 73.2% | xhigh | 24 Jul 2026 | — | Source ↗ | |
| 6 | GPT-6 Astra (medium) OpenAI | 72.8% | medium | 3 Sep 2026 | — | Source ↗ | |
| 7 | GPT-5.6 Sol OpenAI | 72.7% | max | 9 Jul 2026 | — | Source ↗ | |
| 8 | Gemini 3.8 Flash (medium) Google DeepMind | 71.0% | medium | 2 Sep 2026 | — | Source ↗ | |
| 9 | GPT-5.6 Sol (xhigh) OpenAI | 70.7% | xhigh | 9 Jul 2026 | — | Source ↗ | |
| 10 | Claude Fable 5 (xhigh) Anthropic | 69.9% | xhigh | 9 Jun 2026 | — | Source ↗ | |
| 11 | Claude Fable 5 Anthropic | 69.7% | max | 9 Jun 2026 | — | Source ↗ | |
| 12 | GPT-5.6 Terra OpenAI | 69.6% | max | 9 Jul 2026 | — | Source ↗ | |
| 13 | GLM 5.3 Open source Z.ai | 69.0% | max | 14 Aug 2026 | — | Source ↗ | |
| 14 | Claude Opus 5 (medium) Anthropic | 68.9% | medium | 24 Jul 2026 | — | Source ↗ | |
| 15 | Kimi K3 Open source MoonshotAI | 68.5% | max | 16 Jul 2026 | — | Source ↗ | |
| 16 | Grok 4.6 (medium) xAI | 67.5% | medium | 12 Aug 2026 | — | Source ↗ | |
| 17 | GPT-5.6 Luna OpenAI | 67.2% | max | 9 Jul 2026 | — | Source ↗ | |
| 18 | GPT-6 Astra (low) OpenAI | 67.0% | low | 3 Sep 2026 | — | Source ↗ | |
| 19 | GPT-5.5 (xhigh) OpenAI | 67.0% | xhigh | 23 Apr 2026 | — | Source ↗ | |
| 20 | Grok 4.6 (xhigh) xAI | 66.7% | xhigh | 12 Aug 2026 | — | Source ↗ | |
| 21 | Gemini 3.7 Flash (medium) Google DeepMind | 65.5% | medium | 13 Aug 2026 | — | Source ↗ | |
| 22 | Claude Fable 5 (medium) Anthropic | 65.4% | medium | 9 Jun 2026 | — | Source ↗ | |
| 23 | Gemini 3.7 Flash Google | 65.3% | high | 13 Aug 2026 | — | Source ↗ | |
| 24 | Grok 4.6 SpaceXAI | 65.2% | high | 12 Aug 2026 | — | Source ↗ | |
| 25 | GPT-5.5 OpenAI | 64.4% | high | 23 Apr 2026 | — | Source ↗ | |
| 26 | GLM 5.3 Flash Open source Z.ai | 63.4% | max | 20 Aug 2026 | — | Source ↗ | |
| 27 | GPT-5.6 Sol (medium) OpenAI | 61.1% | medium | 9 Jul 2026 | — | Source ↗ | |
| 28 | GPT-5.6 Terra (xhigh) OpenAI | 60.2% | xhigh | 9 Jul 2026 | — | Source ↗ | |
| 29 | Claude Fable 5 (low) Anthropic | 59.6% | low | 9 Jun 2026 | — | Source ↗ | |
| 30 | Claude Opus 4.8 Anthropic | 59.0% | max | 28 May 2026 | — | Source ↗ | |
| 31 | Claude Opus 5 (low) Anthropic | 58.1% | low | 24 Jul 2026 | — | Source ↗ | |
| 32 | Qwen3.8 Max (xhigh) Alibaba | 57.5% | xhigh | 2 Aug 2026 | — | Source ↗ | |
| 33 | GPT-5.6 Luna (xhigh) OpenAI | 56.9% | xhigh | 9 Jul 2026 | — | Source ↗ | |
| 34 | Muse Spark 1.2 (xhigh) Meta AI | 54.9% | xhigh | 5 Aug 2026 | — | Source ↗ | |
| 35 | GPT-5.5 (medium) OpenAI | 54.0% | medium | 23 Apr 2026 | — | Source ↗ | |
| 36 | Claude Sonnet 5 Anthropic | 53.8% | max | 30 Jun 2026 | — | Source ↗ | |
| 37 | Gemini 3.7 Flash (low) Google DeepMind | 53.8% | low | 13 Aug 2026 | — | Source ↗ | |
| 38 | Grok 4.5 SpaceXAI | 53.8% | high | 8 Jul 2026 | — | Source ↗ | |
| 39 | Muse Spark 1.1 Meta | 53.3% | — | 9 Jul 2026 | — | Source ↗ | |
| 40 | GPT-5.4 (xhigh) OpenAI | 51.8% | xhigh | 5 Mar 2026 | — | Source ↗ | |
| 41 | Claude Sonnet 5 (xhigh) Anthropic | 49.7% | xhigh | 30 Jun 2026 | — | Source ↗ | |
| 42 | Gemini 3.6 Flash Google | 46.7% | high | 21 Jul 2026 | — | Source ↗ | |
| 43 | GPT-5.6 Sol (low) OpenAI | 45.4% | low | 9 Jul 2026 | — | Source ↗ | |
| 44 | GLM 5.2 Open source Z.ai | 43.8% | max | 16 Jun 2026 | — | Source ↗ | |
| 45 | Grok 4.6 (low) xAI | 41.6% | low | 12 Aug 2026 | — | Source ↗ | |
| 46 | Claude Sonnet 5 (medium) Anthropic | 39.8% | medium | 30 Jun 2026 | — | Source ↗ | |
| 47 | Gemini 3.5 Flash Google | 37.4% | medium | 19 May 2026 | — | Source ↗ | |
| 48 | GPT-5.6 Terra (medium) OpenAI | 35.1% | medium | 9 Jul 2026 | — | Source ↗ | |
| 49 | Kimi K2.7 Code Open source MoonshotAI | 30.5% | — | 12 Jun 2026 | — | Source ↗ | |
| 50 | Claude Sonnet 5 (low) Anthropic | 30.5% | low | 30 Jun 2026 | — | Source ↗ | |
| 51 | Claude Sonnet 4.6 Anthropic | 29.9% | high | 17 Feb 2026 | — | Source ↗ | |
| 52 | GPT-5.5 (low) OpenAI | 27.0% | low | 23 Apr 2026 | — | Source ↗ | |
| 53 | GPT-5.6 Terra (low) OpenAI | 24.1% | low | 9 Jul 2026 | — | Source ↗ | |
| 54 | Gemini 3.1 Pro Preview Google | 11.7% | — | 19 Feb 2026 | — | Source ↗ | |
| 55 | GPT-5.6 Luna (medium) OpenAI | 11.3% | medium | 9 Jul 2026 | — | Source ↗ | |
| 56 | GPT-5.6 Luna (low) OpenAI | 1.5% | low | 9 Jul 2026 | — | Source ↗ |
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.