On 2026-07-07, the Winzheng YZ Index Smoke Quick Test covered 11 models. Claude Opus 4.7 and Grok 4 tied for first place with a score of 96.99. Smoke is a daily 10-question quick test, suitable for observing short-term signals, and is not equivalent to the Full weekly ranking conclusions.
This Smoke evaluation only covers two main dimensions: Code Execution and Material Constraints. The main ranking formula is 0.55 × Code Execution + 0.45 × Material Constraints. Due to the small daily sample size, single-day scores are more suitable as monitoring signals rather than long-term conclusions about model capabilities.
Daily Rankings
| Rank | Model | Main Score | Code Execution | Material Constraints | Integrity |
|---|---|---|---|---|---|
| #1 | Claude Opus 4.7 | 96.99 | 100 | 93.3 | pass |
| #2 | Grok 4 | 96.99 | 100 | 93.3 | pass |
| #3 | Claude Sonnet 4.6 | 91.44 | 98.5 | 82.8 | pass |
| #4 | GPT-5.5 | 82.77 | 100 | 61.7 | pass |
| #5 | DeepSeek V4 Pro | 80.19 | 98.5 | 57.8 | pass |
| #6 | GPT-o3 | 76.24 | 100 | 47.2 | pass |
| #7 | Gemini 2.5 Pro | 68.94 | 98.5 | 32.8 | pass |
| #8 | Gemini 3.1 Pro | 68.01 | 100 | 28.9 | pass |
| #9 | 豆包 Pro | 67.51 | 100 | 27.8 | pass |
| #10 | Qwen3 Max | 61.34 | 76.1 | 43.3 | pass |
| #11 | GLM-4.6 | 24.26 | 0 | 53.9 | warn |
Data Interpretation
Today's Smoke Quick Test shows that Claude Opus 4.7 and Grok 4 have a main score of 96.99 each, with Code Execution of 100 and Material Constraints of 93.3, forming a structure of perfect Code Execution paired with high Material Constraints; Claude Sonnet 4.6 has a main score of 91.44, Code Execution 98.5, Material Constraints 82.8, with Material Constraints lower than the top two but still maintaining a relatively high main ranking; GPT-5.5 has a main score of 82.77, Code Execution 100, Material Constraints 61.7, showing a combination of perfect Code Execution with weaker Material Constraints.
Claude Opus 4.7's main score increased by 26.8 points, Code Execution +50; Claude Sonnet 4.6 main score +19.9, Code Execution +48.5, and Material Constraints -15; Grok 4 main score +17.6, Code Execution +33.3. These models showed significant improvement in the Code Execution dimension; 豆包 Pro main score -16.4, Material Constraints -67; GLM-4.6 main score -14.5, Code Execution -50, and Integrity changed from pass to warn. As a small-sample single-day signal, the above changes may be due to question sampling fluctuations or actual degradation, and require subsequent runs for verification.
Key Changes
- Claude Opus 4.7: Main score up 26.8 points, Code Execution +50
- Claude Sonnet 4.6: Main score up 19.9 points, Code Execution +48.5, Material Constraints -15
- Grok 4: Main score up 17.6 points, Code Execution +33.3
- 豆包 Pro: Main score down 16.4 points, Code Execution +25, Material Constraints -67
- GLM-4.6: Main score down 14.5 points, Code Execution -50, Material Constraints +28.9, Integrity pass → warn
Signals to Watch
- No abnormal signals worth publishing were retained this time.
When reading this type of Smoke brief, the focus should be on two questions: first, whether a model exposes the same type of weakness for multiple consecutive days; second, whether the integrity rating changes from pass to warn or fail. Large single-day changes in execution or constraint scores may be due to question sampling or could be early signals of actual degradation, requiring subsequent runs for verification.
Data source: YZ Index | Run #218 | View raw data
© 2026 Winzheng.com 赢政天下 | 转载请注明来源并附原文链接