Calibration
AA Coding Index → LiveCodeBench v6
LiveCodeBench v6 is estimated from AA Coding Index with a Michaelis–Menten + offset curve fitted on 9 models measured on both: y = 0.5038 + 0.7610·x / (0.58411 + x), R² = 0.96, cross-validated error 3.1 pp. It is used for 61 estimates.
| Estimated model | AA Coding Index | LiveCodeBench v6 | Source |
|---|---|---|---|
| Apodex 1.1 | 60.8% | 89.2% | estimated ± 3.1 pp, high confidence |
| Apodex 1.1 Mini | 60.8% | 89.2% | estimated ± 3.1 pp, high confidence |
| Celeris-1 | 14.4% | 65.4% | estimated ± 3.1 pp, high confidence |
| Claude 3 Opus | 19.5% | 69.5% | estimated ± 3.1 pp, high confidence |
| Command A+ | 27.9% | 75.0% | estimated ± 3.1 pp, high confidence |
| DeepSeek V3 | 23.0% | 71.9% | estimated ± 3.1 pp, high confidence |
| Gemini 1.5 Pro | 23.6% | 72.3% | estimated ± 3.1 pp, high confidence |
| Gemini 2.5 Pro | 33.3% | 78.0% | estimated ± 3.1 pp, high confidence |
| Gemini 3.1 Pro | 68.8% | 91.5% | estimated ± 3.1 pp, high confidence |
| Gemini 3.6 Flash | 69.2% | 91.7% | estimated ± 3.1 pp, high confidence |
| Gemini 3.7 Flash | 76.1% | 93.4% | estimated ± 3.1 pp, medium confidence |
| Gemini 3.8 Flash | 76.3% | 93.5% | estimated ± 3.1 pp, medium confidence |
| Gemma 3 27B | 10.1% | 61.6% | estimated ± 3.1 pp, high confidence |
| Gemma 4 26B A4B | 39.3% | 81.0% | estimated ± 3.1 pp, high confidence |
| Gemma 4 31B | 43.4% | 82.8% | estimated ± 3.1 pp, high confidence |
| Gemma 4 E2B | 7.2% | 58.8% | estimated ± 3.1 pp, medium confidence |
| Gemma 4 E4B | 9.4% | 60.9% | estimated ± 3.1 pp, high confidence |
| GLM-4.7 | 45.3% | 83.6% | estimated ± 3.1 pp, high confidence |
| GLM-5.3 | 74.8% | 93.1% | estimated ± 3.1 pp, medium confidence |
| GPT-4.1 mini | 20.2% | 69.9% | estimated ± 3.1 pp, high confidence |
| GPT-4.1 nano | 11.1% | 62.6% | estimated ± 3.1 pp, high confidence |
| GPT-4 Turbo | 21.5% | 70.8% | estimated ± 3.1 pp, high confidence |
| GPT-4o mini | 11.4% | 62.8% | estimated ± 3.1 pp, high confidence |
| GPT-5.1 | 49.4% | 85.2% | estimated ± 3.1 pp, high confidence |
| GPT-5.4 mini | 56.1% | 87.7% | estimated ± 3.1 pp, high confidence |
| GPT-5.4 nano | 56.1% | 87.7% | estimated ± 3.1 pp, high confidence |
| GPT-5 (high) | 37.8% | 80.3% | estimated ± 3.1 pp, high confidence |
| GPT-6 Astra | 76.9% | 93.6% | estimated ± 3.1 pp, medium confidence |
| GPT-OSS 120B | 30.4% | 76.5% | estimated ± 3.1 pp, high confidence |
| GPT-OSS 20B | 20.7% | 70.3% | estimated ± 3.1 pp, high confidence |
| Grok 4.3 | 42.3% | 82.3% | estimated ± 3.1 pp, high confidence |
| Grok 4.6 | 76.8% | 93.6% | estimated ± 3.1 pp, medium confidence |
| Hy3 | 58.8% | 88.6% | estimated ± 3.1 pp, high confidence |
| Hy3 Preview | 58.8% | 88.6% | estimated ± 3.1 pp, high confidence |
| K-Exaone | 32.1% | 77.4% | estimated ± 3.1 pp, high confidence |
| Kimi K2.5 (Reasoning) | 46.8% | 84.2% | estimated ± 3.1 pp, high confidence |
| Kimi K2.7 Code | 60.8% | 89.2% | estimated ± 3.1 pp, high confidence |
| Kimi K3 | 76.2% | 93.5% | estimated ± 3.1 pp, medium confidence |
| Ling 2.6 Flash | 25.3% | 73.4% | estimated ± 3.1 pp, high confidence |
| Ling 3.0 Flash FP8 | 50.6% | 85.7% | estimated ± 3.1 pp, high confidence |
| Llama 4 Maverick | 16.3% | 67.0% | estimated ± 3.1 pp, high confidence |
| Llama 4 Scout | 8.2% | 59.7% | estimated ± 3.1 pp, high confidence |
| MiMo-V2-Flash | 49.8% | 85.4% | estimated ± 3.1 pp, high confidence |
| Mistral Large 3 | 20.1% | 69.8% | estimated ± 3.1 pp, high confidence |
| Mistral Medium 3.5 128B | 46.9% | 84.3% | estimated ± 3.1 pp, high confidence |
| Mistral Small 4 | 26.6% | 74.2% | estimated ± 3.1 pp, high confidence |
| Mistral Small 4 (Reasoning) | 26.6% | 74.2% | estimated ± 3.1 pp, high confidence |
| Muse Spark 1.2 | 72.2% | 92.5% | estimated ± 3.1 pp, high confidence |
| Muse Spark 1.3 | 75.8% | 93.4% | estimated ± 3.1 pp, medium confidence |
| Nemotron 3.5 Lightning 30B A3B NVFP4 | 26.8% | 74.3% | estimated ± 3.1 pp, high confidence |
| Nemotron 3 Nano 30B | 14.4% | 65.4% | estimated ± 3.1 pp, high confidence |
| Nemotron 3 Nano Omni 30B A3B | 13.8% | 64.9% | estimated ± 3.1 pp, high confidence |
| Nemotron 3 Super 100B | 37.7% | 80.2% | estimated ± 3.1 pp, high confidence |
| o1 | 39.7% | 81.2% | estimated ± 3.1 pp, high confidence |
| o1-preview | 34.1% | 78.4% | estimated ± 3.1 pp, high confidence |
| Quasar 438B | 61.2% | 89.3% | estimated ± 3.1 pp, high confidence |
| Qwen3.5-122B-A10B | 45.7% | 83.8% | estimated ± 3.1 pp, high confidence |
| Qwen3.8 Max Preview | 71.8% | 92.3% | estimated ± 3.1 pp, high confidence |
| Trinity-Large-Preview | 25.8% | 73.7% | estimated ± 3.1 pp, high confidence |
| Trinity-Large-Thinking | 25.8% | 73.7% | estimated ± 3.1 pp, high confidence |
| Ultravox v0.6 Llama 3.3 70B | 11.9% | 63.3% | estimated ± 3.1 pp, high confidence |