Toxic-comment-detector-AI

Your Fast, Free, AI-Powered Toxic Comment Detector Detecting toxic comment has traditionally been slow and costly. But not anymore. With HateHoundAPI, you can now swiftly identify and filter out toxic content in your web applications. Powered by state-of-the-art AI technology, HateHoundAPI offers:

Demo

https://hate-hound-api.vercel.app/

API Usage

Make predict

  POST /api/predict

parameter	Type	Description
comment	string	Required. The comment you want to predict
access_token	string	Required. You can go to our website, connect your github and get this access_token.

API Input Example

const res = await fetch("https://hate-hound-api.vercel.app/api/predict", {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ comment , access_token}),
})

API Output

HTTP/1.1  201
Location: /api/predict
Content-Type: application/json

{
"comment": "<your-comment>",
"duration": 0.1196323779999986,
"results": {
    "toxic": true,
    "severe_toxic": true,
    "obscene": true,
    "threat": false,
    "insult": true,
    "identity_hate": false
},
"prediction_probs": {
    "toxic": 0.9350948333740234,
    "severe_toxic": 0.4974823296070099,
    "obscene": 0.9486715793609619,
    "threat": 0.13421966135501862,
    "insult": 0.7589452862739563,
    "identity_hate": 0.22608500719070435
}

Support

For support you can just star this project

Licence

MIT