Calibration
AA Coding Index → PostTrainBench v1.1
PostTrainBench v1.1 is estimated from AA Coding Index with a Hill curve fitted on 11 models measured on both: y = 0.0831 + (1.2000 − 0.0831)·x^6.00 / (0.93738^6.00 + x^6.00), R² = 0.55, cross-validated error 6.1 pp. It is used for 46 estimates.
| Estimated model | AA Coding Index | PostTrainBench v1.1 | Source |
|---|---|---|---|
| Apodex 1.1 | 60.8% | 16.0% | estimated ± 6.1 pp, low confidence |
| Apodex 1.1 Mini | 60.8% | 16.0% | estimated ± 6.1 pp, low confidence |
| Celeris-1 | 14.4% | 8.3% | estimated ± 6.1 pp, low confidence |
| Claude 3 Opus | 19.5% | 8.3% | estimated ± 6.1 pp, low confidence |
| Command A+ | 27.9% | 8.4% | estimated ± 6.1 pp, low confidence |
| DeepSeek V3 | 23.0% | 8.3% | estimated ± 6.1 pp, low confidence |
| Gemini 1.5 Pro | 23.6% | 8.3% | estimated ± 6.1 pp, low confidence |
| Gemma 3 27B | 10.1% | 8.3% | estimated ± 6.1 pp, low confidence |
| Gemma 4 12B | 31.0% | 8.5% | estimated ± 6.1 pp, low confidence |
| Gemma 4 26B A4B | 39.3% | 8.9% | estimated ± 6.1 pp, low confidence |
| Gemma 4 31B | 43.4% | 9.4% | estimated ± 6.1 pp, low confidence |
| Gemma 4 E2B | 7.2% | 8.3% | estimated ± 6.1 pp, low confidence |
| Gemma 4 E4B | 9.4% | 8.3% | estimated ± 6.1 pp, low confidence |
| GPT-4.1 mini | 20.2% | 8.3% | estimated ± 6.1 pp, low confidence |
| GPT-4.1 nano | 11.1% | 8.3% | estimated ± 6.1 pp, low confidence |
| GPT-4 Turbo | 21.5% | 8.3% | estimated ± 6.1 pp, low confidence |
| GPT-4o mini | 11.4% | 8.3% | estimated ± 6.1 pp, low confidence |
| GPT-5.1 | 49.4% | 10.7% | estimated ± 6.1 pp, low confidence |
| GPT-5 (high) | 37.8% | 8.8% | estimated ± 6.1 pp, low confidence |
| GPT-OSS 120B | 30.4% | 8.4% | estimated ± 6.1 pp, low confidence |
| GPT-OSS 20B | 20.7% | 8.3% | estimated ± 6.1 pp, low confidence |
| Hy3 | 58.8% | 14.7% | estimated ± 6.1 pp, low confidence |
| Hy3 Preview | 58.8% | 14.7% | estimated ± 6.1 pp, low confidence |
| K-Exaone | 32.1% | 8.5% | estimated ± 6.1 pp, low confidence |
| Kimi K2.5 (Reasoning) | 46.8% | 10.0% | estimated ± 6.1 pp, low confidence |
| LFM2.5-2.6B | 7.7% | 8.3% | estimated ± 6.1 pp, low confidence |
| Ling 2.6 Flash | 25.3% | 8.4% | estimated ± 6.1 pp, low confidence |
| Ling 3.0 Flash FP8 | 50.6% | 11.0% | estimated ± 6.1 pp, low confidence |
| Llama 4 Maverick | 16.3% | 8.3% | estimated ± 6.1 pp, low confidence |
| Llama 4 Scout | 8.2% | 8.3% | estimated ± 6.1 pp, low confidence |
| MiMo-V2-Flash | 49.8% | 10.8% | estimated ± 6.1 pp, low confidence |
| Mistral Large 3 | 20.1% | 8.3% | estimated ± 6.1 pp, low confidence |
| Mistral Small 4 | 26.6% | 8.4% | estimated ± 6.1 pp, low confidence |
| Mistral Small 4 (Reasoning) | 26.6% | 8.4% | estimated ± 6.1 pp, low confidence |
| Nemotron 3.5 Lightning 30B A3B NVFP4 | 26.8% | 8.4% | estimated ± 6.1 pp, low confidence |
| Nemotron 3 Nano 30B | 14.4% | 8.3% | estimated ± 6.1 pp, low confidence |
| Nemotron 3 Nano Omni 30B A3B | 13.8% | 8.3% | estimated ± 6.1 pp, low confidence |
| Nemotron 3 Super 100B | 37.7% | 8.8% | estimated ± 6.1 pp, low confidence |
| o1 | 39.7% | 9.0% | estimated ± 6.1 pp, low confidence |
| o1-preview | 34.1% | 8.6% | estimated ± 6.1 pp, low confidence |
| Quasar 438B | 61.2% | 16.3% | estimated ± 6.1 pp, low confidence |
| Qwen3.5-122B-A10B | 45.7% | 9.8% | estimated ± 6.1 pp, low confidence |
| Qwen3.8 Max Preview | 71.8% | 27.1% | estimated ± 6.1 pp, medium confidence |
| Trinity-Large-Preview | 25.8% | 8.4% | estimated ± 6.1 pp, low confidence |
| Trinity-Large-Thinking | 25.8% | 8.4% | estimated ± 6.1 pp, low confidence |
| Ultravox v0.6 Llama 3.3 70B | 11.9% | 8.3% | estimated ± 6.1 pp, low confidence |