GuardPII GLiNER2 PII Demo
Detect PII from text, DOCX, or PDF files with GLiNER2 plus the high-precision rulebase.
Text input
DOCX or PDF input
Drop File Here
- or -
Click to Upload
Detect PII
Model ID
Threshold
↺
0.05
0.95
Use rulebase gap-filler
Use label descriptions
Value broadcast mode
Document parser
Status
Highlighted text
Entities
Entities
#
⋮
label
⋮
text
⋮
start
⋮
end
⋮
score
⋮
source
⋮
page
⋮
JSON result
Download JSON
Debug / Pipeline Visualization
1. Chunks (text splitting)
▼
2. WhitespaceTokenSplitter Output
▼
3. Token → Prediction Merge (how tokens are joined into entities)
▼
4. Per-Chunk Model Predictions
▼