opus-4.6 vs gpt-5.2-codex
Both models ran the same corpus under the same verifier — cursor and opencode respectively. The harness differs where the names differ, and that is part of what separates them.
opus-4.6 leads by 24.7 points, and the intervals do not overlap.
Where they differ
| Weakness | opus-4.6 | gpt-5.2-codex | Difference |
|---|---|---|---|
| CWE-125Out-of-bounds Read | 65.3% | 41.3% | +24.0 |
| CWE-787Out-of-bounds Write | 76.5% | 41.2% | +35.3 |
| CWE-416Use After Free | 46.2% | 30.8% | +15.4 |
| CWE-908Use of Uninitialized Resource | 33.3% | 22.2% | +11.1 |
| CWE-415Double Free | 66.7% | 50.0% | +16.7 |
| CWE-1284Improper Validation of Specified Quantity in Input | 50.0% | 50.0% | 0.0 |
| CWE-843Access of Resource Using Incompatible Type ('Type Confusion') | 50.0% | 0.0% | +50.0 |
| CWE-476NULL Pointer Dereference | 50.0% | 0.0% | +50.0 |
| CWE-475 | 100.0% | 0.0% | +100.0 |
| CWE-562 | 100.0% | 100.0% | 0.0 |
Per-class samples are small; a difference here is a hint about where to look, not a result. The corpus-level intervals above are the honest comparison.
Corpus cve-bench-136 · run February 2026 · How we test
See the benchmark →