Calibration
GPQA Diamond → ARC-AGI-1
ARC-AGI-1 is estimated from GPQA Diamond with a Hill curve fitted on 12 models measured on both: y = 0.0000 + (1.2000 − 0.0000)·x^6.00 / (0.75439^6.00 + x^6.00), R² = 0.66, cross-validated error 2.7 pp. It is used for 51 estimates.
| Estimated model | GPQA Diamond | ARC-AGI-1 | Source |
|---|---|---|---|
| A.X K2 | 85.6% | 81.7% | estimated ± 2.7 pp, medium confidence |
| Claude Opus 4.6 | 89.2% | 87.9% | estimated ± 2.7 pp, high confidence |
| DeepSeek V4.1 Flash | 90.9% | 90.4% | estimated ± 2.7 pp, high confidence |
| Gemma 4 12B | 78.8% | 67.8% | estimated ± 2.7 pp, medium confidence |
| GLM-5 | 86.0% | 82.4% | estimated ± 2.7 pp, medium confidence |
| GLM-5.1 | 86.2% | 82.8% | estimated ± 2.7 pp, medium confidence |
| GLM-5.2 | 91.2% | 90.9% | estimated ± 2.7 pp, high confidence |
| Hy3 Preview | 87.2% | 84.6% | estimated ± 2.7 pp, medium confidence |
| Hy4 preview | 92.3% | 92.4% | estimated ± 2.7 pp, high confidence |
| Inkling | 87.9% | 85.7% | estimated ± 2.7 pp, medium confidence |
| Interfaze Beta | 89.9% | 88.9% | estimated ± 2.7 pp, high confidence |
| K-EXAONE 2.0 | 82.2% | 75.1% | estimated ± 2.7 pp, medium confidence |
| Kimi K2.6 | 90.5% | 89.9% | estimated ± 2.7 pp, high confidence |
| Kimi K2.5 | 87.6% | 85.2% | estimated ± 2.7 pp, medium confidence |
| LFM2.5-230M | 25.4% | 0.2% | estimated ± 2.7 pp, medium confidence |
| Ling 3.0 Flash | 85.0% | 80.5% | estimated ± 2.7 pp, medium confidence |
| Ling 3.0 Flash FP8 | 84.0% | 78.7% | estimated ± 2.7 pp, medium confidence |
| LLaDA2.2-mini | 44.4% | 4.8% | estimated ± 2.7 pp, medium confidence |
| LongCat-Flash-Lite-Sparse | 69.5% | 45.5% | estimated ± 2.7 pp, medium confidence |
| MAI-Thinking-1 | 84.2% | 79.1% | estimated ± 2.7 pp, medium confidence |
| Mellum2-12B-A2.5B-Instruct | 40.9% | 3.0% | estimated ± 2.7 pp, medium confidence |
| Mellum2-12B-A2.5B-Thinking | 57.6% | 19.8% | estimated ± 2.7 pp, medium confidence |
| Mercury 2.5 | 79.0% | 68.2% | estimated ± 2.7 pp, medium confidence |
| MiniCPM5-1B | 26.3% | 0.2% | estimated ± 2.7 pp, medium confidence |
| MiniCPM5-2B | 70.2% | 47.2% | estimated ± 2.7 pp, medium confidence |
| MiniMax M2.7 | 87.0% | 84.2% | estimated ± 2.7 pp, medium confidence |
| Nemotron 3.5 Lightning 30B A3B NVFP4 | 75.6% | 60.3% | estimated ± 2.7 pp, medium confidence |
| Nemotron 3 Nano Omni 30B A3B | 72.2% | 52.1% | estimated ± 2.7 pp, medium confidence |
| Nemotron 3 Ultra | 87.0% | 84.2% | estimated ± 2.7 pp, medium confidence |
| Ornith-1.5-35B-A3B | 89.2% | 87.9% | estimated ± 2.7 pp, high confidence |
| Ornith-1.5-397B | 92.8% | 93.1% | estimated ± 2.7 pp, high confidence |
| Ornith-1.5-9B | 86.4% | 83.2% | estimated ± 2.7 pp, medium confidence |
| Pareto 26.10 Preview | 92.4% | 92.6% | estimated ± 2.7 pp, high confidence |
| Qwen3.7 Max | 92.4% | 92.6% | estimated ± 2.7 pp, high confidence |
| Qwen3.7 Plus | 90.3% | 89.6% | estimated ± 2.7 pp, high confidence |
| Qwen3.8-27B | 89.2% | 87.9% | estimated ± 2.7 pp, high confidence |
| Qwen3.8-Flash-Next | 91.7% | 91.6% | estimated ± 2.7 pp, high confidence |
| Qwen3.8 Max | 92.6% | 92.9% | estimated ± 2.7 pp, high confidence |
| Qwen3.8-Omni-Flash | 91.0% | 90.6% | estimated ± 2.7 pp, high confidence |
| Beam | 90.5% | 89.9% | estimated ± 2.7 pp, high confidence |
| Sakana Fugu | 95.5% | 96.5% | estimated ± 2.7 pp, high confidence |
| Sakana Fugu-Ultra | 95.5% | 96.5% | estimated ± 2.7 pp, high confidence |
| Solar Open 2 | 86.3% | 83.0% | estimated ± 2.7 pp, medium confidence |
| Solar Pro 4 | 89.0% | 87.5% | estimated ± 2.7 pp, high confidence |
| Soofi S 30B-A3B | 43.4% | 4.2% | estimated ± 2.7 pp, medium confidence |
| Step 5 Preview | 93.5% | 94.1% | estimated ± 2.7 pp, high confidence |
| Ternary Bonsai 2 27B | 85.8% | 82.0% | estimated ± 2.7 pp, medium confidence |
| Trinity-Large-Preview | 63.3% | 31.0% | estimated ± 2.7 pp, medium confidence |
| Trinity-Large-Thinking | 76.3% | 62.0% | estimated ± 2.7 pp, medium confidence |
| ZAYA1-74B-Preview | 57.3% | 19.3% | estimated ± 2.7 pp, medium confidence |
| ZAYA1-8B | 71.0% | 49.2% | estimated ± 2.7 pp, medium confidence |