| # | Model | Score | Relative | Setting | Released | Run | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | Claude Fable 5 (unknown) Anthropic | 53.5% | unknown | 9 Jun 2026 | — | Source ↗ | |
| 2 | Claude Opus 5 Anthropic | 53.4% | max | 24 Jul 2026 | — | Source ↗ | |
| 3 | GPT-6 Astra OpenAI | 53.3% | max | 3 Sep 2026 | — | Source ↗ | |
| 4 | Claude Fable 5.1 (medium) Anthropic | 50.9% | medium | 1 Sep 2026 | — | Source ↗ | |
| 5 | SWE-2 (max) Cognition | 50.0% | max | 10 Sep 2026 | — | Source ↗ | |
| 6 | Grok 4.6 (unknown) xAI | 48.0% | unknown | 12 Aug 2026 | — | Source ↗ | |
| 7 | GPT-5.6 Sol (unknown thinking) OpenAI | 47.5% | unknown | 9 Jul 2026 | — | Source ↗ | |
| 8 | Claude Opus 4.8 Anthropic | 46.5% | unknown | 28 May 2026 | — | Source ↗ | |
| 9 | Kimi K3 (unknown) Open source Moonshot | 44.2% | unknown | 16 Jul 2026 | — | Source ↗ | |
| 10 | Gemini 3.7 Flash (unknown) Google DeepMind | 43.6% | unknown | 13 Aug 2026 | — | Source ↗ | |
| 11 | GPT-5.5 (unknown thinking) OpenAI | 43.0% | unknown | 23 Apr 2026 | — | Source ↗ | |
| 12 | Claude Sonnet 5 (unknown thinking) Anthropic | 42.7% | unknown | 30 Jun 2026 | — | Source ↗ | |
| 13 | Grok 4.5 (unknown thinking) xAI | 42.4% | unknown | 8 Jul 2026 | — | Source ↗ | |
| 14 | SWE-1.7 Cognition | 42.0% | — | 8 Jul 2026 | — | Source ↗ | |
| 15 | GPT-5.6 Terra (unknown thinking) OpenAI | 41.3% | unknown | 9 Jul 2026 | — | Source ↗ | |
| 16 | Gemini 3.8 Flash (medium) Google DeepMind | 41.2% | medium | 2 Sep 2026 | — | Source ↗ | |
| 17 | GLM 5.3 Open source Z.ai | 40.1% | max | 14 Aug 2026 | — | Source ↗ | |
| 18 | GPT-5.6 Luna (unknown thinking) OpenAI | 39.8% | unknown | 9 Jul 2026 | — | Source ↗ | |
| 19 | Claude Opus 4.7 (unknown) Anthropic | 38.5% | unknown | 16 Apr 2026 | — | Source ↗ | |
| 20 | Gemini 3.6 Flash Google | 34.4% | unknown | 21 Jul 2026 | — | Source ↗ | |
| 21 | GLM 5.3 Flash Open source Z.ai | 31.8% | max | 20 Aug 2026 | — | Source ↗ | |
| 22 | Kimi K2.7 Code Open source MoonshotAI | 30.1% | — | 12 Jun 2026 | — | Source ↗ | |
| 23 | DeepSeek V4 Pro 0813 Open source DeepSeek | 28.5% | high | 13 Aug 2026 | — | Source ↗ | |
| 24 | GPT-5.4 mini (unknown thinking) OpenAI | 27.0% | unknown | 17 Mar 2026 | — | Source ↗ | |
| 25 | Claude Opus 4.6 (unknown thinking) Anthropic | 26.6% | unknown | 5 Feb 2026 | — | Source ↗ | |
| 26 | Composer 2.5 Cursor | 25.6% | — | 18 May 2026 | — | Source ↗ | |
| 27 | GLM 5.2 Open source Z.ai | 24.5% | none | 16 Jun 2026 | — | Source ↗ | |
| 28 | Claude Sonnet 4.6 (unknown thinking) Anthropic | 24.3% | unknown | 17 Feb 2026 | — | Source ↗ | |
| 29 | DeepSeek V4 Flash 0731 (unknown) Open source DeepSeek | 18.8% | unknown | 31 Jul 2026 | — | Source ↗ | |
| 30 | DeepSeek v4 Pro (unknown thinking) Open source DeepSeek | 17.6% | none | 24 Apr 2026 | — | Source ↗ | |
| 31 | MiniMax M3 Open source MiniMax | 14.7% | — | 1 Jun 2026 | — | Source ↗ | |
| 32 | Inkling Open source Thinking Machines | 14.0% | — | 15 Jul 2026 | — | Source ↗ | |
| 33 | Nemotron 3 Ultra Open source NVIDIA | 13.6% | — | 4 Jun 2026 | — | Source ↗ | |
| 34 | Qwen3.7 Plus Qwen | 10.2% | — | 2 Jun 2026 | — | Source ↗ | |
| 35 | SWE-1.6 Cognition | 9.4% | — | 7 Apr 2026 | — | Source ↗ | |
| 36 | Mistral Medium 3.5 Open source Mistral | 8.0% | — | 28 Apr 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.