AI benchmark results and model performance
Individual benchmark scores plotted by date.
| Organisation | Model | Reported | Top Score | Info | Self Reported | Source |
|---|---|---|---|---|---|---|
| Ling 2.6 1T | - | 10.80 | 940 puzzles | No | Source | |
| Mistral Medium 3.1 | 12 Aug 2025 | 15.50 | 940 puzzles | No | Source | |
| Mistral Large 3.0 | 02 Dec 2025 | 17.20 | 940 puzzles | No | Source | |
| Llama 4 Maverick | 05 Apr 2025 | 18.40 | 940 puzzles | No | Source | |
| Gemini 3.1 Flash Lite Preview | 03 Mar 2026 | 19.70 | 940 puzzles | No | Source | |
| Ernie 5.0 | 22 Jan 2026 | 21.20 | 940 puzzles | No | Source | |
| MiniMax M2.1 | 23 Dec 2025 | 22.70 | 940 puzzles | No | Source | |
| Qwen 3 Max Thinking | 26 Jan 2026 | 24.20 | 940 puzzles | No | Source | |
| Grok 4.1 Non Thinking | 17 Nov 2025 | 25.10 | Fast non-reasoning; 940 puzzles | No | Source | |
| Mistral Medium 3.5 | 29 Apr 2026 | 25.20 | High; 940 puzzles | No | Source | |
| Claude Haiku 4.5 | 15 Oct 2025 | 26 | 940 puzzles | No | Source | |
| Nemotron 3 Super | 11 Mar 2026 | 26.90 | 940 puzzles | No | Source | |
| MiniMax M2 | 27 Oct 2025 | 27 | 940 puzzles | No | Source | |
| Trinity Large Thinking | 01 Apr 2026 | 29.50 | 940 puzzles | No | Source | |
| MiniMax M2.5 | 12 Feb 2026 | 29.60 | 940 puzzles | No | Source | |
| Hy3 Preview | 23 Apr 2026 | 30.20 | 940 puzzles | No | Source | |
| LongCat Flash Chat | - | 31 | Thinking; 940 puzzles | No | Source | |
| MiniMax M2.7 | 18 Mar 2026 | 35.30 | 940 puzzles | No | Source | |
| Step 3.5 Flash | - | 36.70 | 940 puzzles | No | Source | |
| Claude Opus 4.7 | 16 Apr 2026 | 41 | High reasoning; 940 puzzles | No | Source | |
| MiMo V2 Pro | 18 Mar 2026 | 41 | 940 puzzles | No | Source | |
| Seed 2.0 Pro | 14 Feb 2026 | 42.10 | 940 puzzles | No | Source | |
| Qwen3 Max 2026 01 23 | - | 42.10 | 940 puzzles | No | Source | |
| MiMo V2.5 Pro | 23 Apr 2026 | 48 | 940 puzzles | No | Source | |
| Claude Sonnet 4.5 | 29 Sept 2025 | 49.40 | 16K thinking; 940 puzzles | No | Source |
