PIIMB PII Masking - Avg F1: leaderboard

PIIMB zero-shot PII masking benchmark averaging F1 across OpenPII, Gretel, Nemotron-PII, and Privy.

Metric: Average F1 (%). Source: huggingface.co. 36 models tracked.

Top models

#ModelScoreOverall rank
1OpenMed privacy-filter-multilingual-v292.38
2OpenMed OpenMed-PII-SuperClinical-Large-434M-v187.76
3OpenMed privacy-filter-nemotron-v286.22
4VidaiUK stirling-minimo-pii-10m86.15
5knowledgator gliner-stream-pii-v1.085.96
6antoniogr7 pii-identifier-v0.185.75
7kalyan-ks ettin-68m-nemotron-pii83.56
8OpenMed OpenMed-PII-SuperClinical-Small-44M-v183.27
9gravitee-io bert-small-pii-detection81.7
10kalyan-ks ettin-32m-nemotron-pii81.35

Interactive version: theaggregate.ai/benchmark?slug=piimb-pii-masking-avg-f1 · How It Works · Data refreshed daily, snapshot 2026-10-20.