Calibration
AA Coding Index → NL2Repo
NL2Repo is estimated from AA Coding Index with a linear curve fitted on 13 models measured on both: y = 0.7789·x + -0.0285, R² = 0.73, cross-validated error 5.1 pp. It is used for 51 estimates.
| Estimated model | AA Coding Index | NL2Repo | Source |
|---|---|---|---|
| Apodex 1.1 | 60.8% | 44.5% | estimated ± 5.1 pp, medium confidence |
| Apodex 1.1 Mini | 60.8% | 44.5% | estimated ± 5.1 pp, medium confidence |
| Celeris-1 | 14.4% | 8.4% | estimated ± 5.1 pp, low confidence |
| Claude 3 Opus | 19.5% | 12.4% | estimated ± 5.1 pp, low confidence |
| Claude Fable 5.1 | 81.6% | 60.7% | estimated ± 5.1 pp, low confidence |
| Claude Opus 4.7 (Adaptive) | 73.6% | 54.5% | estimated ± 5.1 pp, medium confidence |
| Command A+ | 27.9% | 18.8% | estimated ± 5.1 pp, low confidence |
| DeepSeek V3 | 23.0% | 15.1% | estimated ± 5.1 pp, low confidence |
| Gemini 1.5 Pro | 23.6% | 15.6% | estimated ± 5.1 pp, low confidence |
| Gemma 3 27B | 10.1% | 5.0% | estimated ± 5.1 pp, low confidence |
| Gemma 4 26B A4B | 39.3% | 27.8% | estimated ± 5.1 pp, low confidence |
| Gemma 4 31B | 43.4% | 31.0% | estimated ± 5.1 pp, medium confidence |
| Gemma 4 E2B | 7.2% | 2.8% | estimated ± 5.1 pp, low confidence |
| Gemma 4 E4B | 9.4% | 4.5% | estimated ± 5.1 pp, low confidence |
| GPT-4.1 mini | 20.2% | 12.9% | estimated ± 5.1 pp, low confidence |
| GPT-4.1 nano | 11.1% | 5.8% | estimated ± 5.1 pp, low confidence |
| GPT-4 Turbo | 21.5% | 13.9% | estimated ± 5.1 pp, low confidence |
| GPT-4o mini | 11.4% | 6.0% | estimated ± 5.1 pp, low confidence |
| GPT-5.1 | 49.4% | 35.6% | estimated ± 5.1 pp, medium confidence |
| GPT-5.4 | 71.1% | 52.5% | estimated ± 5.1 pp, medium confidence |
| GPT-5 (high) | 37.8% | 26.6% | estimated ± 5.1 pp, low confidence |
| GPT-6 Astra | 76.9% | 57.1% | estimated ± 5.1 pp, low confidence |
| GPT-OSS 120B | 30.4% | 20.9% | estimated ± 5.1 pp, low confidence |
| GPT-OSS 20B | 20.7% | 13.3% | estimated ± 5.1 pp, low confidence |
| Hy3 | 58.8% | 43.0% | estimated ± 5.1 pp, medium confidence |
| Hy3 Preview | 58.8% | 43.0% | estimated ± 5.1 pp, medium confidence |
| K-Exaone | 32.1% | 22.2% | estimated ± 5.1 pp, low confidence |
| Kimi K2.5 (Reasoning) | 46.8% | 33.6% | estimated ± 5.1 pp, medium confidence |
| Ling 2.6 Flash | 25.3% | 16.8% | estimated ± 5.1 pp, low confidence |
| Ling 3.0 Flash FP8 | 50.6% | 36.6% | estimated ± 5.1 pp, medium confidence |
| Llama 4 Maverick | 16.3% | 9.8% | estimated ± 5.1 pp, low confidence |
| Llama 4 Scout | 8.2% | 3.5% | estimated ± 5.1 pp, low confidence |
| MiMo-V2-Flash | 49.8% | 36.0% | estimated ± 5.1 pp, medium confidence |
| Mistral Large 3 | 20.1% | 12.8% | estimated ± 5.1 pp, low confidence |
| Mistral Small 4 | 26.6% | 17.9% | estimated ± 5.1 pp, low confidence |
| Mistral Small 4 (Reasoning) | 26.6% | 17.9% | estimated ± 5.1 pp, low confidence |
| Muse Glimmer 30B | 49.0% | 35.3% | estimated ± 5.1 pp, medium confidence |
| Muse Spark 1.3 | 75.8% | 56.2% | estimated ± 5.1 pp, low confidence |
| Nemotron 3.5 Lightning 30B A3B NVFP4 | 26.8% | 18.0% | estimated ± 5.1 pp, low confidence |
| Nemotron 3 Nano 30B | 14.4% | 8.3% | estimated ± 5.1 pp, low confidence |
| Nemotron 3 Nano Omni 30B A3B | 13.8% | 7.9% | estimated ± 5.1 pp, low confidence |
| Nemotron 3 Super 100B | 37.7% | 26.5% | estimated ± 5.1 pp, low confidence |
| o1 | 39.7% | 28.1% | estimated ± 5.1 pp, low confidence |
| o1-preview | 34.1% | 23.7% | estimated ± 5.1 pp, low confidence |
| Quasar 438B | 61.2% | 44.8% | estimated ± 5.1 pp, medium confidence |
| Qwen3.5-122B-A10B | 45.7% | 32.8% | estimated ± 5.1 pp, medium confidence |
| Qwen3.8 Max Preview | 71.8% | 53.1% | estimated ± 5.1 pp, medium confidence |
| Step 3.7 Flash | 39.6% | 28.0% | estimated ± 5.1 pp, low confidence |
| Trinity-Large-Preview | 25.8% | 17.2% | estimated ± 5.1 pp, low confidence |
| Trinity-Large-Thinking | 25.8% | 17.2% | estimated ± 5.1 pp, low confidence |
| Ultravox v0.6 Llama 3.3 70B | 11.9% | 6.4% | estimated ± 5.1 pp, low confidence |