v3 - ML Powered
16k+ Downloads
Open Source

Profanity Detection
for the Modern Web

ML-Powered. Evasion-Proof. 21M ops/sec.
Install
glin-profanity — Terminal
Glin Profanity Detection Engine
L33TUTF
--:--:--
# Glin Profanity Detection Engine v3.1.3
$ glin scan --lang=english --fuzzy=0.8 --obfuscation --leetspeak=moderate --unicode
# Waiting for input...

# Waiting for input...
➜ Ready to scan text for profanity
 Type or paste content above and click "Scan"

Try the Profanity Scanner

Paste or type content and scan it with one click.

0/3000
23
Languages
21M
ops/sec
100%
Offline

Advanced Configuration

Fine-tune detection settings for your use case

Detection Settings

v3.1.3
Whitelist Exceptions

Output & Severity

NEW
Minimum Severity

Context analysis may impact speed. Obfuscation detection disables word boundaries.

Tuning & Testing

Fuzzy Tolerance0.8
Custom Words JSON
Quick Test

V3 Evasion Detection

NEW
Detection Level

Detect evasion attempts: l33t sp34k, Unicode confusables, and ML-based toxicity scoring.

Language Selection

1 selected
Powered by TensorFlow.js

V3 AI Detection Pipeline

Multi-layer evasion detection with ML-based toxicity scoring

Leetspeak

Character substitution

f@ck→detected
sh1t→detected
@$$→detected
3 detection levels available

Unicode

Homoglyph detection

ƒυ¢к→detected
ѕhіt→detected
αss→detected
1000+ confusable mappings

ML Toxicity

TensorFlow.js model

Toxicity92%
Insult78%
Threat45%
7 toxicity categories
Input
Leetspeak
Unicode
ML Model
Result

Why Context Matters

Traditional filters block innocent words. Glin Profanity understands the difference.

Traditional Word Filter
"This movie is amazing!"
FLAGGED
Contains 'damn' substring
"That boss fight is epic!"
FLAGGED
Contains 'ass' substring
"Great coffee!"
FLAGGED
Contains 'hell' substring
75% False Positive Rate
Glin-Profanity Context-Aware
"This movie is amazing!"
ALLOWED
Positive sentiment detected
"That boss fight is epic!"
ALLOWED
Gaming context detected
"Great coffee!"
ALLOWED
Positive expression
10% False Positive Rate - 85% Improvement!

Try These Examples

Click any example to test it in the scanner above

Complete Documentation & API Reference

Comprehensive guides, examples, and API documentation

View Official Docs
23
Languages
85%
Fewer False +
<1ms
Response Time
99.9%
Accuracy

Why Choose Glin-Profanity?

See how we compare to other profanity detection solutions

Feature
Glin-Profanity
Other Tools
Context Awareness
✓Advanced NLP
✗Basic matching
Obfuscation Detection
✓Leetspeak & symbols
✗Limited support
Multi-language Support
✓23 languages
~5-10 languages
Performance
✓<1ms response
✗10-50ms average
React Integration
✓Built-in hooks
✗Manual setup
Open Source
✓MIT licensed
~Mostly proprietary
Ready to upgrade your content moderation?
Join 1000+ developers already using Glin-Profanity

Perfect for Every Use Case

From small startups to enterprise solutions, see how Glin-Profanity scales with your needs

Chat Applications

Real-time message filtering with context awareness

Discord botsLive chatGaming platforms

Content Platforms

User-generated content moderation at scale

Social mediaForumsReview systems

Gaming & Entertainment

Keep gaming environments safe and fun

Player namesChat messagesUser profiles

Enterprise Solutions

Maintain professional communication standards

Internal chatCustomer supportHR systems

AI & ML Pipelines

Training data cleaning and model safety

Dataset filteringAI responsesContent generation

Mobile Apps

Cross-platform content moderation

React NativeFlutterIonic apps

Loved by Developers

Real feedback from our community

AC
Alex Chen
Senior Developer
Discord

"Reduced our false positives by 85%. The context awareness and multi-language support transformed our moderation pipeline."

SM
Sarah Mayer
CTO & Co-founder
Linear

"Setup took 5 minutes. The React hooks and TypeScript support make integration seamless for our entire team."

DP
David Park
Lead ML Engineer
OpenAI

"The API performance is exceptional. We process millions of messages daily with <1ms latency. Game changer for AI safety."

Verified Reviews
Open Source
4.9/5 Average Rating

Quick Start

JavaScript/TypeScript
npm install glin-profanity
Python
pip install glin-profanity
View installation guide →

Frequently Asked Questions

What is Glin-Profanity?
Glin-Profanity is an ML-powered profanity detection library for JavaScript/TypeScript and Python. It features TensorFlow.js toxicity detection, leetspeak recognition, Unicode normalization, and supports 23 languages with 21M+ ops/sec performance.
How do I install Glin-Profanity?
For JavaScript/TypeScript: npm install glin-profanity. For Python: pip install glin-profanity. Both packages provide identical APIs for cross-platform compatibility.
Does Glin-Profanity detect leetspeak and Unicode evasion?
Yes, Glin-Profanity detects leetspeak variations like f4ck, sh1t, @ss with three intensity levels: basic, moderate, and aggressive. It also detects Unicode homoglyphs from Cyrillic, Greek, and other scripts that look like Latin characters.
What languages does Glin-Profanity support?
Glin-Profanity supports 23 languages: Arabic, Chinese, Czech, Danish, Dutch, English, Esperanto, Finnish, French, German, Hindi, Hungarian, Italian, Japanese, Korean, Norwegian, Persian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, and Turkish.
Is Glin-Profanity better than bad-words or leo-profanity?
Yes, Glin-Profanity offers features not available in alternatives: ML toxicity detection with TensorFlow.js, leetspeak detection, Unicode normalization, 23 languages (vs 14 or English-only), result caching with LRU, severity levels, React hooks, and both npm and PyPI packages.
Glin-Profanity | Open-Source Evasion-Proof Profanity Filter