Claude Opus 4.7 Leads with 78.72: 2026-11-23 Smoke Quick Test Data Brief

On 2026-11-23, the YZ Index Smoke quick test covered 14 models, with Claude Opus 4.7 ranking first for the day at 78.72. Smoke is a daily 10-question quick test, suitable for observing short-term signals and not equivalent to conclusions from the Full weekly leaderboard.

This Smoke evaluation covers only two main leaderboard dimensions: code execution and material constraints. The main leaderboard formula is 0.55 × code execution + 0.45 × material constraints. Because the daily sample size is small, single-day scores are more suitable as monitoring signals than as long-term conclusions about model capability.

Daily Ranking

RankModelMainCode ExecutionMaterial ConstraintsIntegrity
#1Claude Opus 4.778.729657.6pass
#2GPT-6.1 Sol78.2999.352.6pass
#3Claude Sonnet 4.677.998.652.6pass
#4GPT-6 Sol75.2799.345.9pass
#5GPT-5.566.7974.357.6pass
#6Grok 466.3577.652.6warn
#7GPT-6 Astra64.5474.352.6pass
#8GPT-6 Luna64.2796.425pass
#9Doubao Pro62.0173.847.6pass
#10GPT-o358.2873.839.3pass
#11DeepSeek V4 Pro56.445064.3pass
#12GLM-4.656.445064.3warn
#13Gemini 3.1 Pro51.175052.6pass
#14Gemini 2.5 Pro48.165045.9pass

Data Interpretation

In today's YZ Index Smoke quick test, leading models showed clear divergence in the balance between code execution and material constraints. Claude Opus 4.7 ranked first with a main leaderboard score of 78.72, with code execution at 96 and material constraints at 57.6; GPT-6.1 Sol had a main leaderboard score of 78.29, with code execution at 99.3 and material constraints at 52.6; Claude Sonnet 4.6 had a main leaderboard score of 77.9, with code execution at 98.6 and material constraints at 52.6. These models generally scored higher on code execution than on material constraints, forming a structure of strong code execution paired with moderate material constraints. GPT-6 Sol had a main leaderboard score of 75.27, with code execution at 99.3 and material constraints at 45.9, also reflecting this pattern.

Compared with the previous run under the same methodology, some models showed significant score changes. Claude Sonnet 4.6 rose 16.3 points on the main leaderboard, with code execution up 50 points and material constraints down 24.8 points; GPT-5.5 rose 12.6 points on the main leaderboard, with code execution up 25 points; GLM-4.6 rose 12.7 points on the main leaderboard, with code execution up 25 points. GPT-o3 fell 20.1 points on the main leaderboard, with material constraints down 47.4 points; Grok 4 fell 17.2 points on the main leaderboard, with code execution down 10.9 points and material constraints down 24.8 points, and its integrity rating moved from pass to warn; GPT-6 Luna's material constraints fell 36.9 points; Doubao Pro fell 11.2 points on the main leaderboard; Gemini 2.5 Pro fell 9.4 points on the main leaderboard.

Anomalous signals include the sharp drop in Claude Sonnet 4.6's material constraints, the sharp drop in Grok 4's main leaderboard score, the sharp drop in GPT-6 Luna's material constraints, the sharp drop in Doubao Pro's main leaderboard score, the sharp drop in GPT-o3's main leaderboard score, and the sharp drop in Gemini 2.5 Pro's main leaderboard score. These changes may stem from question sampling fluctuation or may reflect genuine degradation; follow-up runs are needed to confirm. As a small-sample single-day signal, the current Smoke data is for reference only and does not constitute a basis for long-term judgment.

Key Changes

  • GPT-o3: Main leaderboard down 20.1 points; material constraints -47.4 points
  • Grok 4: Main leaderboard down 17.2 points; code execution -10.9 points; material constraints -24.8 points; integrity pass→warn
  • Claude Sonnet 4.6: Main leaderboard up 16.3 points; code execution +50 points; material constraints -24.8 points
  • GLM-4.6: Main leaderboard up 12.7 points; code execution +25 points
  • GPT-5.5: Main leaderboard up 12.6 points; code execution +25 points

Signals to Watch

  • Claude Sonnet 4.6: Material constraints plunged -24.8 points
  • Grok 4: Main leaderboard plunged -17.2 points
  • GPT-6 Luna: Material constraints plunged -36.9 points
  • Doubao Pro: Main leaderboard plunged -11.2 points
  • GPT-o3: Main leaderboard plunged -20.1 points
  • Gemini 2.5 Pro: Main leaderboard plunged -9.4 points
  • Qwen3 Max: Incomplete data (several evaluation dimensions missing due to API failure/timeout); automatically queued for rerun and not ranked in this issue

When reading this type of Smoke brief, focus on two questions: first, whether a model has consistently exposed the same type of weakness over multiple days; second, whether its integrity rating has moved from pass to warn or fail. Large single-day changes in execution or constraint scores may come from question sampling or may be early signals of genuine degradation, requiring follow-up runs for verification.


Data source: YZ Index | Run #439 | View raw data

This article is from Winzheng Index blog, translated in full by Winzheng (winzheng.com). Click here to view the original When republishing the translation, please credit the source. Thank you!