NLP API

Toxic Content Detection API

Deteksi ujaran kebencian, cyberbullying, kata kasar, dan konten toxic secara otomatis untuk menjaga kualitas interaksi di aplikasi, forum, komunitas, dan platform digital.

Toxicity Detection

Mengidentifikasi konten toxic, kata kasar, dan perilaku tidak pantas dalam teks.

Hate Speech Detection

Membantu mendeteksi ujaran kebencian dan diskriminasi secara otomatis.

Real-Time Moderation

Cocok untuk moderasi komentar, chat, forum, dan platform komunitas.

🟢 Endpoint

POST https://kubus.co.id/api/v1/moderation/toxic/batch

Enable

🔐 Headers

Use your API key to authorize requests. Replace YOUR_API_KEY with your actual key.

                Authorization: Bearer YOUR_API_KEY
                Content-Type: application/json

Manage Your API Keys

📦 Request Body

Configured Entity Groups

No entity groups configured yet.

Example request body:


            {
                "entity_group": "toxic_1",
                "texts": [
                    "You are stupid",
                    "I love this!",
                    "That was the worst service ever",
                    "Great work team!"
                ]
            }

✅ Response

A successful response will identify whether each text is toxic, and if so, the category:


            {
            "results": [
                {
                "category": "insult",
                "text": "You are stupid",
                "toxic": true
                },
                {
                "text": "I love this!",
                "toxic": false
                },
                {
                "category": "harsh_criticism",
                "text": "That was the worst service ever",
                "toxic": true
                },
                {
                "text": "Great work team!",
                "toxic": false
                }
            ],
            "status": "success"
            }

Notes

toxic: Indicates if the content is considered harmful.
category: Specifies the type of toxicity (e.g., insult, hate speech, harsh criticism).
Maximum 10 texts per batch.
Use this API to moderate user-generated content in forums, chats, reviews, etc.