APIs & SDKs
Unprofane
Profanity filtering your users can't get around
An API that detects profanity and toxicity in 17 languages, catching disguises like leetspeak and look-alike letters while avoiding false positives on places
Overview
A look at unprofane.com
Unprofane is a profanity and toxicity detection API designed to catch profanity and abuse across 17 languages. It detects disguised profanity including leetspeak, look-alike letters from other alphabets, misspellings, spoonerisms and puns. The service distinguishes between swearing in general and swearing directed at someone, applying different policies (kids, standard, casual) to appropriately handle context.
The API is positioned as a successor to Google's Perspective API, which ends December 31, 2026, with compatible JSON and error formats. It offers a rules-based engine for most texts with an optional AI model for complex cases, and operates in fast mode (rules only) or balanced mode (model-assisted). Unprofane publishes open benchmarks showing it catches 89% of toxic texts at the same false-alarm rate where Perspective caught 58.4%, and maintains a public "Scunthorpe Test" to verify it avoids blocking innocent place names and surnames.
Feature notes
- Catches leetspeak, look-alike letters and creative spelling variants
- Context-aware filtering with three policies: kids, standard, casual
- Detects abuse without explicit swear words
- Avoids false positives on place names, surnames and anatomy
- Supports 17 languages
- Drop-in replacement for Google Perspective API with compatible JSON and errors
- Rules-based engine with optional AI model for complex cases
- Fast mode (rules only) and balanced mode (model-assisted)
- Open benchmark showing performance metrics
- Available for username checks, chat, comments, reviews and AI agent responses
Pricing
Pricing model: Paid
Hobby: $9/month (25,000 checks); Starter: $29/month (100,000 checks); Growth: $249/month (1,000,000 checks); Scale: $999/month (5,000,000 checks). Extra checks available at $29 per 100,000 (Hobby/Starter), $25 per 100,000 (Growth), $20 per 100,000 (Scale). Launch offer: half price on Hobby and Start
Compare and explore
FAQ
- What counts as a check?
- One text item sent to the check endpoint, the batch endpoint (each item counts), the Perspective endpoint or the MCP server. A request refused for a bad body, a bad key or a rate limit does not count. A cached answer counts because the work of keeping the cache is the same. An account's test keys share one cap of 100 checks a day on every plan and without one.
- What happens when I cancel?
- The plan runs to the end of the period you paid for. After that the account has no plan: live and publishable keys stop, test keys keep working, and nothing is deleted. Choosing a plan again brings the live keys back.
- What if it gets a text wrong?
- Report a miss through the dashboard's 'Wrong result' button or via email with the check id. Every correction is answered and you are notified when it is fixed.
- Is there yearly billing?
- Growth and Scale can be paid a year ahead: twelve months for the price of ten, with twelve months of checks to use across the year. Hobby and Starter are monthly. Enterprise can be invoiced monthly or yearly.
Leave a review for Unprofane
No reviews of Unprofane yet. Run the first inspection.
Reviews are free and are never paid for. Open the reviews
Reviews and notes · Change log · Share the profile · Maker of this product? Claim the profile · Report a fault
Unprofane has a free listing. The owner can unlock the full profile with a paid launch for $5:
- Founder story (locked)
- Milestones (locked)
- Demo video (locked)
- Dofollow website link (locked)
Comparable tools
- Kinetic.placeDR 0 · 0 community votes
- Reddit Data APIDR 0 · 0 community votes
- localmemDR 0 · 0 community votes
- IPProxyDR 0 · 0 community votes
