Text Moderation made easy

Moderate and filter user texts in real-time with our API. Protect your users. Protect your brand.

Have questions? Talk to sales

woman's profile photo

Hi! My name's Jane and I love traveling

Safe. Processing took 8ms

man's profile photo

What the f&$ck happened?

Profanity found. Processing took 6ms

man's profile photo

Like s_*_x? I'm at rick(at)gmail(dot)com

Sexual content and Personal Email found.

woman's profile photo

I really enjoyed this book. Took me less than 2 weeks to read!

Safe. Processing took 12ms

female profile photo

I ignore j33rks like him

Profanity found. Processing took 9ms

Automatically find inappropriate words and phrases in any text

Our Text Moderation API has been designed and battle-tested on all sorts of text inputs. It is available in many languages so that you can support users from any region. Use it to moderate user messages, chats, user profiles, usernames, comments or reviews in real-time.

Messages, live chats, forums

woman's profile photo
male profile photo

Long-form reviews & posts

woman's profile photo

Short messages & comments

man's profile photo

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 its position in the text, so you can easily remove the offending phrase, change it or block the entire message.

  • Insults and Swear words
  • Sexual content
  • Racist and Discriminatory language
  • Inappropriate language
  • Grawlix
  • mizzpelingz
  • l33t spe4|<
  • s̼P̑̇̈e⃣c̸i͢aⓁ characters
  • reeppppettitiioonns
  • I--N...S   E R,,..T_I * O°N°S
  • uʍop ǝpᴉsdn

Prevent users from bypassing your filters

Our smart filters will work even when users actively try to circumvent them.

Using Natural Language Analysis and Proprietary Algorithms, our filters are able to detect millions of variations for each unwanted word: letter replacements, character insertions, phonetic variations, leet speak, obfuscation, hyphenation, special characters, embeddings etc.

Detect Personal Information

Automatically detect email addresses, phone numbers from multiple countries, IP addresses, social security numbers and more.

Our email detection filters work even when users actively try to bypass your filters, such as with jane(at)gmail(dot)com or robert_hotmail_com.

Email address

U.K. phone number

IPv4 address

U.S. phone number

Email address

IPv6 address

Canadian phone number

Social Security Number

Protect your users— Prevent users from sharing personal details when they shouldn't or when they might have been tricked into sharing. Protect their privacy.

Compliance— Make sure you don't display or store Personally Identifiable Information (PII) when you should not. Make sure your are GDPR-compliant.

Detect links and URLs

Decide when users should be allowed to share links and web URLs. Prevent spammers from abusing your platform and block unwanted ads or scams containing links.

You can implement your own whitelist to allow specific URLs, for instance to allow only URLs pointing to your own domains. You can also implement a selective blacklist to block any URL pointing to your competitors.

All things considered I would rather http://shadysite.co/buy8 or you could click on bit.ly/e3rJLad to get some Last month I could spend some time with https://allowedsite.com is whitelisted

Add your own custom blacklists and whitelists

Easily add any word. Filter competitor's names or words specific to your audience.

All options are Highly configurable.

Benefit from our smart match engine. Any word you add will match with millions of variations, while still ignoring false positives.

Extremely fast and efficient

Your customers expect their content to be validated correctly and immediately — not eventually.
We constantly monitor our accuracy and speed.
21ms median text
processing time

Instant response. No callback mechanisms for Text Moderation. No queues. No waiting.

Optimized for speed. Powerful Hardware for Text Analysis.

5+ regions to
choose from

Geo-distribution. For enterprise customers: ability to choose the region processing your requests, both for speed optimization and for compliance.

Full scalability. Process anywhere from a single request to billions per month.

GET STARTED NOW

Quick and Easy to Integrate

Our API was designed by developers for developers. Copy paste a few lines of code and you will be ready to go! Access our detailed documentation, quick start guides or talk to our live support at anytime.
    
    
# copy and paste this in your terminal
curl -X POST 'https://api.sightengine.com/1.0/text/check.json' \
  -F 'text=Contact rick123(at)gmail(dot)com to have s_*_x' \
  -F 'lang=en' \
  -F 'mode=rules,ml'
SEE THE DOCUMENTATION SIGN UP

We also do Image and Video Moderation

We are the one-stop-shop for all your content moderation needs. Moderate photos and videos with our easy-to-use APIs.

Standard Moderation

Nudity & Adult content

Detect nudity across 29 classes from very explicit to mildly suggestive

Violence

Detect physical violence and threats

Weapons

Detect weapons, firearms, threatening knives

Recreational & Medical Drugs

Detect pills, cannabis, joints, syringes, heroin...

Alcoholic beverages

Detect wine, beer, cocktails, spirits

Self-Harm

Detect self-harm

Hate & offensive signs

Detect hate, offensive, nazi, supremacist signs and gestures

Gore & Disgusting

Detect horrific content, blood, corpses, wounds...

Smoking & Tobacco Products

Detect smoking and tobacco products

Gambling

Detect gambling, casinos, slot machines...

Money & Banknotes

Detect displays of money and banknotes

Face / People related

Minors

Detect babies, children and teenagers under 18

Sunglasses

Detect faces hidden with sun glasses

Text analysis, QR codes and OCR

Text Moderation in Images/Videos

Detect profanity and personal information

Embedded texts

Detect texts and watermarks

QR code Moderation

Moderate QR codes

AI Generation and Editing

AI-generated image detection

Detect if image was AI-generated

Deepfake detection

Detect deepfakes

Content Description

Image quality

Detect images that have a low quality such as bad lighting, excessive blur, bad contrast

Color detection

Detect the dominant and secondary colors

Image type detection

Determine if an image is a natural photo or an illustration

Near-duplicate Detection

Image near-duplicate detection

Detect similar and nearly identical images

Video near-duplicate detection

Detect videos containing common scenes

Fraud Detection

Scammers

Block scammers and fraudulent users

Audio

Audio profanity detection

Determine if a video contains spoken profanity or insults

Ready to get a taste of the future?

GET STARTED NOW