| abstract-reasoning-on-arc-agi | Claude Fable 5 (Max) | #1 | Accuracy: 98.5 |
| abstract-reasoning-on-arc-agi | Gemini 3.8 Flash (High) | #2 | Accuracy: 98.5 |
| abstract-reasoning-on-arc-agi | Claude Opus 5.5 (High) | #4 | Accuracy: 98.5 |
| abstract-reasoning-on-arc-agi | Gemini 4 Argon (High) | #5 | Accuracy: 98.5 |
| abstract-reasoning-on-arc-agi | GPT-5.6 Sol (XHigh) | #6 | Accuracy: 97.5 |
| abstract-reasoning-on-arc-agi | Kimi K3 (Max) | #8 | Accuracy: 94.5 |
| abstract-reasoning-on-arc-agi | DeepSeek V4.1 Flash (Max) | #9 | Accuracy: 94.5 |
| abstract-reasoning-on-arc-agi | GLM-5.3-Flash (Max) | #10 | Accuracy: 91.0 |
| abstract-reasoning-on-arc-agi | DeepSeek V4 Flash 0731 (Max) | #12 | Accuracy: 89.0 |
| abstract-reasoning-on-arc-agi | Inkling | #14 | Accuracy: 79.5 |
| abstract-reasoning-on-arc-agi | Kimi K2.5 | #15 | Accuracy: 65.33 |
| abstract-reasoning-on-arc-agi | DeepSeek V3.2 | #16 | Accuracy: 57 |
| abstract-reasoning-on-arc-agi | Tiny Recursion Model (TRM) | #18 | Accuracy: 40.0 |
| abstract-reasoning-on-arc-agi | Deepseek R1 (05/28) | #19 | Accuracy: 21.21 |
| abstract-reasoning-on-arc-agi | Qwen3-235b-a22b Instruct (25/07) | #20 | Accuracy: 11.0 |
| abstract-reasoning-on-arc-agi | Llama 4 Maverick | #21 | Accuracy: 4.38 |
| abstract-reasoning-on-arc-agi | Claude 3.7 (16K) | – | Accuracy: 28.6 |
| abstract-reasoning-on-arc-agi | Claude 4.7 (High) | – | Accuracy: 93.5 |
| abstract-reasoning-on-arc-agi | Claude Fable 5.1 (Max) | – | Accuracy: 97.5 |
| abstract-reasoning-on-arc-agi | Claude Haiku 5.5 (Max) | – | Accuracy: 89.0 |
| abstract-reasoning-on-arc-agi | Claude Opus 4.6 (120K, High) | – | Accuracy: 94.0 |
| abstract-reasoning-on-arc-agi | Claude Opus 4.8 (Max) | – | Accuracy: 92.5 |
| abstract-reasoning-on-arc-agi | Claude Sonnet 4 (Thinking 16K) | – | Accuracy: 40.0 |
| abstract-reasoning-on-arc-agi | Claude Sonnet 4.5 (Thinking 32K) | – | Accuracy: 63.67 |
| abstract-reasoning-on-arc-agi | GPT-5 Pro | – | Accuracy: 70.17 |
| abstract-reasoning-on-arc-agi | GPT-5.5 Pro (High) | – | Accuracy: 96.5 |
| abstract-reasoning-on-arc-agi | GPT-6 Sol (Max) | – | Accuracy: 95.5 |
| abstract-reasoning-on-arc-agi | GPT-6.1 Sol (XHigh) | – | Accuracy: 98.5 |
| abstract-reasoning-on-arc-agi | Gemini 3 Deep Think (2/26) | – | Accuracy: 96.0 |
| abstract-reasoning-on-arc-agi | Gemini 3.1 Pro (Preview) | – | Accuracy: 98.0 |
| abstract-reasoning-on-arc-agi | Gemini 3.5 Flash (High) | – | Accuracy: 92.5 |
| abstract-reasoning-on-arc-agi | Gemini 3.6 Flash (High) | – | Accuracy: 91.167 |
| abstract-reasoning-on-arc-agi | Grok 4.6 (Medium) | – | Accuracy: 87.5 |
| abstract-reasoning-on-arc-agi | Magistral Medium (Thinking) | – | Accuracy: 6.12 |
| abstract-reasoning-on-arc-agi-2 | Claude Opus 5.5 (High) | #2 | Accuracy: 93.3333 |
| abstract-reasoning-on-arc-agi-2 | GPT-5.6 Sol (Max) | #3 | Accuracy: 92.5 |
| abstract-reasoning-on-arc-agi-2 | Gemini 4 Argon (High) | #4 | Accuracy: 91.25 |
| abstract-reasoning-on-arc-agi-2 | Claude Fable 5 (Max) | #6 | Accuracy: 89.167 |
| abstract-reasoning-on-arc-agi-2 | Gemini 3.8 Flash (High) | #7 | Accuracy: 89.1667 |
| abstract-reasoning-on-arc-agi-2 | DeepSeek V4.1 Flash (Max) | #8 | Accuracy: 72.9167 |
| abstract-reasoning-on-arc-agi-2 | GLM-5.3-Flash (Max) | #9 | Accuracy: 65.8333 |
| abstract-reasoning-on-arc-agi-2 | DeepSeek V4 Flash 0731 (Max) | #10 | Accuracy: 61.389 |
| abstract-reasoning-on-arc-agi-2 | Kimi K3 (Max) | #11 | Accuracy: 60.417 |
| abstract-reasoning-on-arc-agi-2 | Inkling | #14 | Accuracy: 36.528 |
| abstract-reasoning-on-arc-agi-2 | GLM-5.2 | #15 | Accuracy: 22.7750 |
| abstract-reasoning-on-arc-agi-2 | Kimi K2.5 | #16 | Accuracy: 11.81 |
| abstract-reasoning-on-arc-agi-2 | o3 (High) | #17 | Accuracy: 6.53 |
| abstract-reasoning-on-arc-agi-2 | Tiny Recursion Model (TRM) | #18 | Accuracy: 6.25 |
| abstract-reasoning-on-arc-agi-2 | MiniMax M2.5 | #20 | Accuracy: 4.86 |
| abstract-reasoning-on-arc-agi-2 | Deepseek V3.2 | #21 | Accuracy: 4.03 |
| abstract-reasoning-on-arc-agi-2 | Llama 4 Maverick | #22 | Accuracy: 0.0 |
| abstract-reasoning-on-arc-agi-2 | ARChitects | – | Accuracy: 2.5 |
| abstract-reasoning-on-arc-agi-2 | Claude 3.7 (8K) | – | Accuracy: 0.9 |
| abstract-reasoning-on-arc-agi-2 | Claude 4.7 (Max) | – | Accuracy: 75.83 |
| abstract-reasoning-on-arc-agi-2 | Claude Fable 5.1 (Max) | – | Accuracy: 90.0 |
| abstract-reasoning-on-arc-agi-2 | Claude Haiku 5.5 (XHigh) | – | Accuracy: 61.2500 |
| abstract-reasoning-on-arc-agi-2 | Claude Opus 4.6 (120K, High) | – | Accuracy: 69.17 |
| abstract-reasoning-on-arc-agi-2 | Claude Opus 4.8 (High) | – | Accuracy: 72.08 |
| abstract-reasoning-on-arc-agi-2 | Claude Sonnet 4.6 (High) | – | Accuracy: 60.42 |
| abstract-reasoning-on-arc-agi-2 | GPT-5.4 Pro (XHigh) | – | Accuracy: 83.33 |
| abstract-reasoning-on-arc-agi-2 | GPT-6 Sol (Max) | – | Accuracy: 89.5833 |
| abstract-reasoning-on-arc-agi-2 | GPT-6.1 Sol (Max) | – | Accuracy: 94.1667 |
| abstract-reasoning-on-arc-agi-2 | Gemini 2.5 Pro (Thinking 32K) | – | Accuracy: 4.86 |
| abstract-reasoning-on-arc-agi-2 | Gemini 3 Deep Think (2/26) | – | Accuracy: 84.58 |
| abstract-reasoning-on-arc-agi-2 | Gemini 3.5 Flash (High) | – | Accuracy: 72.08 |
| abstract-reasoning-on-arc-agi-2 | Gemini 3.6 Flash (High) | – | Accuracy: 60.417 |
| abstract-reasoning-on-arc-agi-2 | Grok 4.5 (High) | – | Accuracy: 52.64 |
| abstract-reasoning-on-arc-agi-2 | Grok 4.6 (XHigh) | – | Accuracy: 67.0833 |
| abstract-reasoning-on-arc-agi-2 | Magistral Medium (Thinking) | – | Accuracy: 0.0 |
| abstract-reasoning-on-arc-agi-2 | NVARC | – | Accuracy: 27.64 |