THE ARCHIVE / SEPTEMBER 4, 2026
Earlier benchmark
results.
Security and correctness-coverage boards for JavaScript, Python, Go, Java, and TypeScript. F1 followed by TP / FP / FN; separate corpora from the newer C++ results.
Vantage Gate · security
Vantage vs CodeQL vs SonarQube · original edition’s Pan-confirmed, un-blinded board · now leads the Security page
| Corpus | Vantage | CodeQL | SonarQube |
|---|---|---|---|
| JS+Python (58) | 0.4894 23/13/35 | 0.1818 6/2/52 | 0.0667 2/0/56 |
| Go | 0.9130 42/8/0 | 0.1333 3/0/39 | 0 genuine zero, 251 ncloc |
| Java | 1.0 52/0/0 | 0.40 14/4/38 | 0.4928 17/0/35 |
| TypeScript | 1.0 48/0/0 | 0.1852 5/1/43 | 0.1786 5/3/43 |
Java circularity disclosure: self-built corpus; CodeQL serves as an independent tool check. This is not a third-party corpus result.
Vantage Code · correctness coverage
Security-tool coverage of a correctness corpus. Low incumbent scores measure incidental coverage, not general tool quality.
| Corpus | Vantage | CodeQL | SonarQube |
|---|---|---|---|
| JS+Python (67) | 0.9041 66/13/1 | 0.2195 9/6/58 | 0.1333 5/3/62 |
| Go (33) | 0.8919† 33/8/0 | 0.1143 2/0/31 | 0.0588 1/0/32 |
| Java (33) | 0.9429† 33/4/0 | 0.0930 2/8/31 | 0.1143 2/0/31 |
| TypeScript (38) | 0.6526 31/26/7 | 0.186 4/1/34 | 0.1 2/0/36 |
† GT-informed development: detectors built with the frozen answer key open, tuned on scorer FN/FP detail; zero GT edits, no GT IDs in engine code. The disclosure travels with the number. The newer, separate C++ correctness board is a decisive loss.