The easiest toxicity-detection API

A simple API to help you flag and censor toxic comments on your platform.

Documentation

Fast

The toxicity model runs in under 100ms.

Accurate

Within 1% accuracy of the top model for the Kaggle Toxic Comment Classification Challenge.

Unbiased

The model has been optimized with a metric that measures unintended bias.

No credit card required

Free tier lets you try out the API without committing to anything.

Password-less authentication

Any account is only as secure as your email, why not reduce the possible points of failure by one.

Easy docs

Documentation built to get you up and running as fast as possible.

Dead-simple integration

Sign up

No payment needed

Get your auth key

Simple as it gets, one key

Use API

Copy and paste right from the docs

Get Started

Checkout the documentation

Documentation