Calibration
SWE-bench Verified → SciCode
SciCode is estimated from SWE-bench Verified with a Michaelis–Menten curve fitted on 14 models measured on both: y = 2.0000·x / (2.49806 + x), R² = 0.81, cross-validated error 3.8 pp. It is used for 38 estimates.
| Estimated model | SWE-bench Verified | SciCode | Source |
|---|---|---|---|
| BTL-4 | 78.4% | 47.8% | estimated ± 3.8 pp, high confidence |
| Claude 3.5 Sonnet | 49.0% | 32.8% | estimated ± 3.8 pp, high confidence |
| Claude 4.1 Opus | 74.5% | 45.9% | estimated ± 3.8 pp, high confidence |
| Claude 4 Sonnet | 72.7% | 45.1% | estimated ± 3.8 pp, high confidence |
| Claude Mythos 5 | 95.5% | 55.3% | estimated ± 3.8 pp, medium confidence |
| Claude Opus 4.5 | 80.9% | 48.9% | estimated ± 3.8 pp, high confidence |
| Claude Opus 4.6 | 80.8% | 48.9% | estimated ± 3.8 pp, high confidence |
| Claude Opus 4.7 (Adaptive) | 87.6% | 51.9% | estimated ± 3.8 pp, medium confidence |
| Claude Sonnet 4.5 | 77.2% | 47.2% | estimated ± 3.8 pp, high confidence |
| dots3-note Preview | 78.4% | 47.8% | estimated ± 3.8 pp, high confidence |
| Ember-1 | 92.2% | 53.9% | estimated ± 3.8 pp, medium confidence |
| GLM-5 | 77.8% | 47.5% | estimated ± 3.8 pp, high confidence |
| GPT-4.1 | 54.6% | 35.9% | estimated ± 3.8 pp, high confidence |
| GPT-4.1 mini | 23.6% | 17.3% | estimated ± 3.8 pp, medium confidence |
| GPT-5.2 | 80.0% | 48.5% | estimated ± 3.8 pp, high confidence |
| Kimi K2.5 (Reasoning) | 76.8% | 47.0% | estimated ± 3.8 pp, high confidence |
| Laguna XS 2.1 | 70.9% | 44.2% | estimated ± 3.8 pp, high confidence |
| LLaDA2.2-flash | 49.3% | 33.0% | estimated ± 3.8 pp, high confidence |
| LongCat-Flash-Lite-Sparse | 68.2% | 42.9% | estimated ± 3.8 pp, high confidence |
| MAI-Code-1.1-Flash | 72.6% | 45.0% | estimated ± 3.8 pp, high confidence |
| MAI-Thinking-1 | 73.5% | 45.5% | estimated ± 3.8 pp, high confidence |
| MiMo-V2-Flash | 73.4% | 45.4% | estimated ± 3.8 pp, high confidence |
| MiMo-V2-Omni | 74.8% | 46.1% | estimated ± 3.8 pp, high confidence |
| MiMo-V2-Pro | 78.0% | 47.6% | estimated ± 3.8 pp, high confidence |
| Muse Spark | 77.4% | 47.3% | estimated ± 3.8 pp, high confidence |
| o3-mini | 49.3% | 33.0% | estimated ± 3.8 pp, high confidence |
| Ornith-1.0-35B | 75.6% | 46.5% | estimated ± 3.8 pp, high confidence |
| Ornith-1.0-397B | 82.4% | 49.6% | estimated ± 3.8 pp, medium confidence |
| Ornith-1.0-9B | 69.4% | 43.5% | estimated ± 3.8 pp, high confidence |
| Ornith-1.5-35B-A3B | 79.0% | 48.1% | estimated ± 3.8 pp, high confidence |
| Ornith-1.5-397B | 86.0% | 51.2% | estimated ± 3.8 pp, medium confidence |
| Ornith-1.5-9B | 70.6% | 44.1% | estimated ± 3.8 pp, high confidence |
| Qwen3.5-27B | 72.4% | 44.9% | estimated ± 3.8 pp, high confidence |
| Qwen3.5-35B-A3B | 69.2% | 43.4% | estimated ± 3.8 pp, high confidence |
| Qwen3.5 397B | 76.2% | 46.7% | estimated ± 3.8 pp, high confidence |
| Solar Open 2 | 70.4% | 44.0% | estimated ± 3.8 pp, high confidence |
| Ternary Bonsai 2 27B | 60.8% | 39.1% | estimated ± 3.8 pp, high confidence |
| ZAYA1-74B-Preview | 53.2% | 35.1% | estimated ± 3.8 pp, high confidence |