EN · PII
tonicai PII detection benchmark results
A frozen evaluation slice with 1,500 rows, 2,417 normalized gold annotations and source-level provenance.
1,500input rows
2,417gold annotations
259,595characters
41measured detectors
Dataset provenance
- Language group
- EN
- Task
- PII
- License
- cc-by-4.0
- Data categories
- ACCOUNT, CONTACT, ORG, PERSON
- Project generated
- No
- Corrupted copy
- No
Rows without annotations are not verified clean. Reported extra masking is not a human-confirmed false-positive rate.
Detector outcomes
Sorted by untouched annotations. Training-source overlaps are excluded.
| Detector | Family | Untouched ↓ | Fully hidden ↑ | Extra masking ↓ | Char F1 ↑ |
|---|---|---|---|---|---|
| gliner2-fastino | GLINER2 | 4.47% | 92.18% | 2.79% | 0.885 |
| gliner2-large | GLINER2 | 5.63% | 91.64% | 5.65% | 0.844 |
| gliner2-hivetrace-omni | GLINER2 | 7.07% | 89.33% | 2.08% | 0.891 |
| gliner-nvidia | GLINER | 7.07% | 88.75% | 1.45% | 0.915 |
| gliner-pii-edge | GLINER | 11.42% | 80.47% | 2.86% | 0.836 |
| gravitee-small | HF | 13.78% | 82.25% | 3.70% | 0.788 |
| gliner-multi-v21 | GLINER | 17.87% | 66.16% | 2.43% | 0.761 |
| pplx | PPLX | 20.77% | 75.51% | 0.75% | 0.848 |
| gliner2-vladlinv | GLINER2 | 23.58% | 74.27% | 0.42% | 0.859 |
| gliner-stream-pii | GLINER | 25.90% | 59.62% | 5.36% | 0.670 |
| gliner25-fastino | GLINER2 | 26.02% | 72.45% | 2.65% | 0.791 |
| mmbert32k | HF | 26.15% | 62.10% | 0.40% | 0.830 |
| kalyan-ettin | HF | 27.22% | 69.01% | 1.28% | 0.816 |
| nuner-zero | GLINER | 28.51% | 67.48% | 5.26% | 0.794 |
| openmed-nemotron | HF | 29.00% | 67.31% | 0.75% | 0.818 |
| bardsai-eu | ONNX | 30.53% | 53.58% | 0.25% | 0.732 |
| gliner2-hivetrace-uni | GLINER2 | 31.24% | 67.44% | 1.23% | 0.814 |
| openai-base | OPF | 31.49% | 65.37% | 0.30% | 0.841 |
| openmed-multilingual | HF | 31.53% | 63.43% | 1.37% | 0.815 |
| nym-small | ONNX | 35.25% | 62.02% | 0.28% | 0.849 |
| ner-ru-yqelz | HF | 35.37% | 57.34% | 5.25% | 0.514 |
| opf-ru | OPF | 36.86% | 59.45% | 0.27% | 0.816 |
| apararti | OPF | 37.28% | 59.33% | 0.36% | 0.815 |
| davlan-xlmr | HF | 39.84% | 55.73% | 0.17% | 0.549 |
| nym-base | HF | 41.50% | 55.44% | 0.40% | 0.816 |
| opf-kz-ru | OPF | 43.40% | 53.37% | 0.21% | 0.794 |
| traciora | OPF | 47.95% | 48.99% | 0.23% | 0.783 |
| ru-pii-ner | RUPII | 48.74% | 48.90% | 0.88% | 0.737 |
| ner-ru-gherman | HF | 50.14% | 38.85% | 0.08% | 0.493 |
| stanza-ru | STANZA | 54.99% | 44.60% | 3.36% | 0.275 |
| davlan-mbert | HF | 55.32% | 39.47% | 0.05% | 0.477 |
| gliner-pii-base | GLINER | 57.72% | 39.35% | 2.29% | 0.560 |
| gliner-urchade | GLINER | 57.80% | 37.73% | 2.93% | 0.670 |
| opf-ru-v2 | OPF | 58.42% | 38.39% | 0.02% | 0.748 |
| ru-legal-ner | HF | 68.14% | 25.16% | 3.74% | 0.554 |
| pii-shield-onnx | ONNX | 72.16% | 25.69% | 0.56% | 0.652 |
| rules-ru | RULES | 80.84% | 15.56% | 0.00% | 0.651 |
| fef2-secret-ru | HF | 85.77% | 10.92% | 0.06% | 0.484 |
| spacy-ru-lg | SPACY | 85.89% | 14.11% | 0.64% | 0.154 |
| natasha | NATASHA | 87.01% | 11.25% | <0.01% | 0.186 |
| spacy-alrosait | SPACY | 99.96% | 0.04% | 0.00% | 0.000 |