Calibration
AA Coding Index → Vibe Code Bench
Vibe Code Bench is estimated from AA Coding Index with a inverse Michaelis–Menten curve fitted on 16 models measured on both: y = 0.22036·(x − 0.0000) / (0.0000 + 0.9934 − x), R² = 0.77, cross-validated error 10.3 pp. It is used for 55 estimates.
| Estimated model | AA Coding Index | Vibe Code Bench | Source |
|---|---|---|---|
| Apodex 1.1 Mini | 60.8% | 34.7% | estimated ± 10.3 pp, low confidence |
| Celeris-1 | 14.4% | 3.7% | estimated ± 10.3 pp, low confidence |
| Claude 3 Opus | 19.5% | 5.4% | estimated ± 10.3 pp, low confidence |
| Claude Fable 5.1 | 81.6% | 100.0% | estimated ± 10.3 pp, low confidence |
| Command A+ | 27.9% | 8.6% | estimated ± 10.3 pp, low confidence |
| Gemini 1.5 Pro | 23.6% | 6.9% | estimated ± 10.3 pp, low confidence |
| Gemini 3.5 Flash-Lite | 49.3% | 21.7% | estimated ± 10.3 pp, low confidence |
| Gemini 3.6 Flash | 69.2% | 50.7% | estimated ± 10.3 pp, low confidence |
| Gemini 3.8 Flash | 76.3% | 72.9% | estimated ± 10.3 pp, low confidence |
| Gemma 3 27B | 10.1% | 2.5% | estimated ± 10.3 pp, low confidence |
| Gemma 4 12B | 31.0% | 10.0% | estimated ± 10.3 pp, low confidence |
| Gemma 4 26B A4B | 39.3% | 14.4% | estimated ± 10.3 pp, low confidence |
| Gemma 4 E2B | 7.2% | 1.7% | estimated ± 10.3 pp, low confidence |
| Gemma 4 E4B | 9.4% | 2.3% | estimated ± 10.3 pp, low confidence |
| GLM-5.2 | 68.8% | 49.5% | estimated ± 10.3 pp, low confidence |
| GLM-5.3 | 74.8% | 67.0% | estimated ± 10.3 pp, low confidence |
| GPT-4.1 nano | 11.1% | 2.8% | estimated ± 10.3 pp, low confidence |
| GPT-4 Turbo | 21.5% | 6.1% | estimated ± 10.3 pp, low confidence |
| GPT-4o mini | 11.4% | 2.9% | estimated ± 10.3 pp, low confidence |
| GPT-5.6 Luna | 71.5% | 56.5% | estimated ± 10.3 pp, low confidence |
| GPT-5.6 Sol | 77.4% | 77.7% | estimated ± 10.3 pp, low confidence |
| GPT-5.6 Terra | 76.7% | 74.5% | estimated ± 10.3 pp, low confidence |
| Grok 4.3 | 42.3% | 16.3% | estimated ± 10.3 pp, low confidence |
| Grok 4.5 | 72.5% | 59.4% | estimated ± 10.3 pp, low confidence |
| Grok 4.6 | 76.8% | 75.0% | estimated ± 10.3 pp, low confidence |
| Hy3 | 58.8% | 32.0% | estimated ± 10.3 pp, low confidence |
| K-Exaone | 32.1% | 10.5% | estimated ± 10.3 pp, low confidence |
| Kimi K2.7 Code | 60.8% | 34.7% | estimated ± 10.3 pp, low confidence |
| Kimi K3 | 76.2% | 72.7% | estimated ± 10.3 pp, low confidence |
| LFM2.5-2.6B | 7.7% | 1.9% | estimated ± 10.3 pp, low confidence |
| Ling 2.6 Flash | 25.3% | 7.5% | estimated ± 10.3 pp, low confidence |
| Ling 3.0 Flash | 50.6% | 22.9% | estimated ± 10.3 pp, low confidence |
| Ling 3.0 Flash FP8 | 50.6% | 22.9% | estimated ± 10.3 pp, low confidence |
| Llama 4 Maverick | 16.3% | 4.3% | estimated ± 10.3 pp, low confidence |
| Llama 4 Scout | 8.2% | 2.0% | estimated ± 10.3 pp, low confidence |
| MiMo-V2.5-Pro | 60.2% | 33.9% | estimated ± 10.3 pp, low confidence |
| Mistral Large 3 | 20.1% | 5.6% | estimated ± 10.3 pp, low confidence |
| Mistral Small 4 | 26.6% | 8.1% | estimated ± 10.3 pp, low confidence |
| Mistral Small 4 (Reasoning) | 26.6% | 8.1% | estimated ± 10.3 pp, low confidence |
| Muse Spark 1.1 | 71.3% | 56.1% | estimated ± 10.3 pp, low confidence |
| Muse Spark 1.2 | 72.2% | 58.7% | estimated ± 10.3 pp, low confidence |
| Muse Spark 1.3 | 75.8% | 70.9% | estimated ± 10.3 pp, low confidence |
| Nemotron 3 Nano 30B | 14.4% | 3.7% | estimated ± 10.3 pp, low confidence |
| Nemotron 3 Nano Omni 30B A3B | 13.8% | 3.5% | estimated ± 10.3 pp, low confidence |
| Nemotron 3 Super 100B | 37.7% | 13.5% | estimated ± 10.3 pp, low confidence |
| o1 | 39.7% | 14.7% | estimated ± 10.3 pp, low confidence |
| o1-preview | 34.1% | 11.5% | estimated ± 10.3 pp, low confidence |
| Quasar 438B | 61.2% | 35.4% | estimated ± 10.3 pp, low confidence |
| Qwen3.8-27B | 68.1% | 48.0% | estimated ± 10.3 pp, low confidence |
| Qwen3.8-Flash-Next | 73.1% | 61.2% | estimated ± 10.3 pp, low confidence |
| Qwen3.8 Max Preview | 71.8% | 57.5% | estimated ± 10.3 pp, low confidence |
| Step 3.7 Flash | 39.6% | 14.6% | estimated ± 10.3 pp, low confidence |
| Trinity-Large-Preview | 25.8% | 7.7% | estimated ± 10.3 pp, low confidence |
| Trinity-Large-Thinking | 25.8% | 7.7% | estimated ± 10.3 pp, low confidence |
| Ultravox v0.6 Llama 3.3 70B | 11.9% | 3.0% | estimated ± 10.3 pp, low confidence |