评测来源/CyberBench Patch v1.1
观察中编程
CyberBench Patch v1.1
Vals AI · CyberBench Patch v1.1 的公开模型评测结果。
本来源的模型成绩
| GPT-6 Sol | 77.98 | 78.0 | 2026-11-05 |
| Gemini 4 Argon | 77.86 | 77.9 | 2026-11-05 |
| GPT-5.6 Sol | 76.31 | 76.3 | 2026-11-05 |
| GPT-6 Luna | 76.25 | 76.3 | 2026-11-05 |
| Claude Haiku 5.5 | 75.48 | 75.5 | 2026-11-05 |
| MiMo V2.6 Flash | 75.36 | 75.4 | 2026-11-05 |
| Kimi K3 | 75.24 | 75.2 | 2026-11-05 |
| DeepSeek V4.1 Flash | 73.69 | 73.7 | 2026-11-05 |
| GPT-5.6 Luna | 73.63 | 73.6 | 2026-11-05 |
| DeepSeek V4 Pro 0813 | 72.86 | 72.9 | 2026-11-05 |
| MiMo V2.6 Pro | 72.86 | 72.9 | 2026-11-05 |
| Muse Spark 1.3 | 72.74 | 72.7 | 2026-11-05 |
| GPT-5.6 Terra | 72.08 | 72.1 | 2026-11-05 |
| DeepSeek V4 Flash 0731 | 72.08 | 72.1 | 2026-11-05 |
| GLM 5.3 | 72.08 | 72.1 | 2026-11-05 |
| Grok 4.5 | 71.01 | 71.0 | 2026-11-05 |
| Claude Fable 5.1 | 70.42 | 70.4 | 2026-11-05 |
| Muse Spark 1.2 | 69.52 | 69.5 | 2026-11-05 |
| Grok 4.7 | 69.46 | 69.5 | 2026-11-05 |
| GLM-5.3 Flash | 67.86 | 67.9 | 2026-11-05 |
| Grok 4.6 | 67.68 | 67.7 | 2026-11-05 |
| Inkling | 66.79 | 66.8 | 2026-11-05 |
| Claude Opus 5 | 65.36 | 65.4 | 2026-11-05 |
| MiniMax-M3 | 65 | 65.0 | 2026-11-05 |
| Inkling Small | 64.11 | 64.1 | 2026-11-05 |
| Claude Sonnet 5 | 61.91 | 61.9 | 2026-11-05 |
| Claude Sonnet 5.5 | 59.58 | 59.6 | 2026-11-05 |
| Gemini 3.5 Flash Lite | 59.11 | 59.1 | 2026-11-05 |
| Claude Opus 5.5 | 55.36 | 55.4 | 2026-11-05 |
| Gemini 3.6 Flash | 45.36 | 45.4 | 2026-11-05 |
| Gemini 3.8 Flash | 43.75 | 43.8 | 2026-11-05 |
| GPT-6 Astra | 41.07 | 41.1 | 2026-11-05 |
| GPT-6.1 Sol | 39.29 | 39.3 | 2026-11-05 |
| Qwen3.8 Max | 28.57 | 28.6 | 2026-11-05 |