Code Migration — Code Quality
Code · 2026-06-16
Code-quality rubric score (0–5) sub-track of the Code Migration benchmark. Split out because it is a 0–5 rubric, not a 0–100 pass rate.
Top models (higher is better)
| Model | Score |
|---|---|
| Sonnet 5 | 4.9 |
| Claude Opus 5 | 4.8 |
| Opus 4.8 | 4.7 |
| Opus 4.7 | 4.6 |
| Claude Fable 5 | 4.4 |
| Grok 4.5 | 4.1 |
| Sonnet 4.6 | 4.0 |
| GLM-5.2 | 4.0 |
| GPT-5.6 Sol | 3.8 |
| MiniMax M3 | 3.7 |
| GPT-5.4 | 3.6 |
| Kimi K2.7 Code | 3.5 |
| GLM-5.1 | 3.5 |
| Qwen3.6 Plus (2026-04-02) | 3.5 |
| GPT-5.5 | 3.5 |
| DeepSeek-V4-Pro | 3.4 |
| MiMo-V2.5-Pro | 3.4 |
| Haiku 4.5 | 3.3 |
| GPT-5.6 Terra | 3.3 |
| Gemini 3.6 Flash | 3.3 |
| GPT-5.6 Luna | 3.2 |
| Kimi K2.6 | 3.1 |
| Grok 4.3 | 3.1 |
| GPT-5.4 Mini | 3.1 |
| MiniMax M2.7 | 3.0 |
| Muse Spark 1.1 | 2.8 |
| Gemini 3.5 Flash | 2.8 |
| MiMo-V2.5 | 2.8 |
| Gemini 3.1 Flash-Lite Preview | 2.7 |
| Gemini 3 Flash Preview | 2.7 |
| Gemini 3.5 Flash-Lite | 2.7 |
| Gemini 3.1 Pro Preview | 2.5 |
| Inkling | 2.2 |