评测来源/Vibe Code Bench 1–100
观察中编程
Vibe Code Bench 1–100
Vals AI · Vibe Code Bench 1–100 的公开模型评测结果。
本来源的模型成绩
| Claude Opus 5.5 | 30.36 | 30.4 | 2026-11-05 |
| Claude Opus 5 | 28.53 | 28.5 | 2026-11-05 |
| Claude Fable 5.1 | 28 | 28.0 | 2026-11-05 |
| GPT-6 Astra | 27.64 | 27.6 | 2026-11-05 |
| GPT-5.6 Luna | 22.59 | 22.6 | 2026-11-05 |
| Gemini 4 Argon | 21.2 | 21.2 | 2026-11-05 |
| Muse Spark 1.3 | 20.46 | 20.5 | 2026-11-05 |
| Claude Haiku 5.5 | 20.41 | 20.4 | 2026-11-05 |
| GPT-5.6 Sol | 20.01 | 20.0 | 2026-11-05 |
| GLM 5.3 | 19.99 | 20.0 | 2026-11-05 |
| Gemini 3.8 Flash | 18.77 | 18.8 | 2026-11-05 |
| Kimi K3 | 18.24 | 18.2 | 2026-11-05 |
| DeepSeek V4 Pro 0813 | 17.53 | 17.5 | 2026-11-05 |
| DeepSeek V4.1 Flash | 16.38 | 16.4 | 2026-11-05 |
| GLM-5.3 Flash | 16.03 | 16.0 | 2026-11-05 |
| GPT-5.6 Terra | 14.82 | 14.8 | 2026-11-05 |
| Grok 4.6 | 14.75 | 14.8 | 2026-11-05 |
| Mistral Large 4 | 14.26 | 14.3 | 2026-10-22 |
| Claude Sonnet 5 | 13.82 | 13.8 | 2026-11-05 |
| Qwen3.8 Max | 12.83 | 12.8 | 2026-11-05 |
| MiniMax-M3 | 9.22 | 9.2 | 2026-11-05 |
| Inkling | 7.33 | 7.3 | 2026-11-05 |