LEAKS detector
gitleaks PII detection benchmark
Measured on 9 of 41 retained datasets under the same frozen scoring and masking protocol.
0.51%fully hidden annotations
0.59%detected / overlapped
0.45%extra masking
—CPU characters / second
Strengths
Highest complete-masking rates by normalized data type.
Passwords, keys & tokens17.36%170 / 979
Logins & usernames0.00%0 / 626
Bank accounts & cards0.00%0 / 1,415
Watch list
Lowest observed complete-masking rates; inspect before deployment.
Other sensitive attributes0.00%0 / 224
Customer & employee IDs0.00%0 / 483
Network identifiers0.00%0 / 1,658
Language coverage
The overall leader can differ from the best choice for one language slice.
| Language | Datasets | Fully hidden ↑ | Detected ↑ | Extra masking ↓ |
|---|---|---|---|---|
| English | 2 | 39.95% | 47.00% | 0.77% |
| Russian | 7 | 0.05% | 0.05% | 0.00% |
Configuration, source and version
- Family
- LEAKS
- Revision
- 8.30.1
- Eligible datasets
- 9 / 41
- Character F1
- 0.132
- Untouched diagnostic
- 99.41% · 33,436 / 33,633
- Benchmark version
- 1.0.3 · 2026-09-09
- CPU seconds / 10k chars
- Not measured
- GPU seconds / 10k chars
- Not measured
source overlap: secrets-rules
Results by dataset
Complete masking is primary; detection shows whether any character overlapped.
| Dataset | Gold annotations | Fully hidden ↑ | Detected ↑ | Extra masking ↓ |
|---|---|---|---|---|
| alrosait | 1,862 | 0.00% | 0.00% | 0.00% |
| jayguard | 1,195 | 0.00% | 0.00% | 0.00% |
| nym-ru | 9,497 | 0.00% | 0.00% | — |
| redmadrobot | 5,516 | 0.00% | 0.00% | 0.00% |
| russian-pii-66k | 4,805 | 0.00% | 0.00% | — |
| scanpatch | 8,708 | 0.00% | 0.00% | 0.00% |
| hivetrace | 1,667 | 1.02% | 1.02% | 0.00% |
| leaky-repo | 95 | 23.16% | 23.16% | 0.00% |
| secrets-issues | 288 | 45.49% | 54.86% | 0.81% |