Dimension Drop
Severity 10/10
2026-W34
DeepSeek V4 Pro Code Execution (v5) Dropped 27.4 pts
Score Comparison
| Dimension | Previous | Current | Change |
|---|---|---|---|
| Overall (v5) | 72.5 | 65.2 | -7.3 |
| Code Execution (v5) | 80.3 | 52.9 | -27.4 |
| Knowledge Synthesis (v5) | 85.6 | 81.7 | -3.9 |
| Grounding (v5) | 71.9 | 76.7 | +4.8 |
| Value | 44.3 | 43.0 | -1.3 |
| Stability | 37.4 | 47.9 | +10.5 |
| Availability | 99.0 | 88.0 | -11 |
Affected Dimensions
Code Execution (v5)
Run #282 · Formula v7 · Judge v6.4 · Benchmark v7 · 2026-08-17 05:04 SGT
View DeepSeek V4 Pro Full Profile