Text Moderation

Moderate and filter user texts in real-time with our API. Keep users happy and engaged by blocking profanities and inappropriate text messages with CyberPurify’s foul language filter.

Automated models with a
human-level understanding of textual content

Analyze and return probability scores that text contains concepts such as toxic, insult, obscene and threat. Moderate content and ensure the exclusion of profanity, threatening language and other unwanted text.

Hi! My name’s Jeane and I love reading.

Safe. Processing took 6ms

What the f&$ck happened?

Profanity found. Processing took 6ms

Send me a pic of your b***s? I’m at jack(at)gmail(dot)com

Sexual content and Personal Email found.

I ignore j33rks like him.

Profanity found. Processing took 6ms

You’re ugly!

Bullying found.

You should kill yourself.

Personal Attacks found.

Effective near-instant removal of profanities and bad words

Increase retention with automated profanity and bad word filters. 85% of users are unlikely to return to your site if they encounter inappropriate content. Clean your user generated content in real time and before it goes live.

Using Natural Language Analysis and Algorithms, our filters are able to detect millions of variations for each unwanted word: spelling mistakes, leet speak, obfuscation, sdditional spacing, Special character insertion, hyphenation etc.

Filter any Objectionable content. You are in control.

Pick and choose the filters you want to apply. No extra work or list creation needed on your end. You can use our filters straight out of the box.

Whenever there is a match, our API gives you a tag describing the type of content found along with it’s position in the text, so you can easily remove the offending phrase, change it or block the entire message.

Add your own custom blacklists and whitelists

Easily add any word. Create new, update existing and import your pre-existing lists to expand on what the filters catch. Customize the setup so it fits your specific needs.

 

We support various input files

TEXT STRINGS

Up to 1,024 characters, supporting five languages: English, French, Vietnamese.

TEXT IN IMAGES (OCR MODEL)

Extract text in images, including memes, GIFs, and screenshots.

We detect content across multiple model classes

Sexual

Classify text for sexually explicit or suggestive content

Violence

Flag content containing extreme threats of violence

Cyberbullying

Identify bullying or abusive language in real time

Hate

Detect hate speech with high levels of granularity

Profanity Filter

Profane words and some phrases in supported languages, with the potential for additional phrases

Custom

Custom user-provided text filters that get applied at each request, managed by users in the API dashboard

Protect your users

Ideal for Any Experience that Allows User-Generated Content

Blogs, Forums & Communities

Don’t let abusive or harmful social interactions compromise the user experience in forums, dating apps, private chat, or live streams.

Customer Reviews

Remove unnecessarily offensive and hateful reviews from your retail, travel, or hospitality sites—and showcase your brand in the best possible light.
w

SMS & In-App Messaging

Moderate in-app chat among staff and customers, live-streaming text campaigns, and SMS content for social walls and more.

Children’s Sites

Protect kids from profanity and other potentially harmful content, and filter PII like email and phone numbers to help support COPPA-compliance.

In-Game Chat

Ensure your players can express themselves freely while maintaining a healthy, inclusive community that drives engagement.

Customer Communications

Protect customers from disgruntled employees—and employees from combative customers—whether it’s via live chat, email, or other.

Integration Model

The Text Moderation offers multiple integration options to suit your needs.

Cloud

Use CyberPurify Text Moderation in the Cloud to reduce IT costs and to speed up deployment.

On-Premise

CyberPurify will help you deploy the Text Moderation on your private servers for full compliance with the privacy regulations.
w