Real-time content moderation, powered by AI.
Detect spam, abuse and harmful content in milliseconds. One API for text and images, with custom rules and detailed audit reports.
Text and image, one API
Classify user-generated content across categories in a single call.
Text moderation
Classify messages, comments, reviews and chat against a rich set of risk categories.
- Spam and promotional content
- Insults and harassment
- Political and sensitive topics
- Adult / NSFW language
- Advertising and link detection
- Custom keyword dictionaries
Image moderation
Scan user uploads and shared media for unsafe content before they reach your users.
- NSFW / adult imagery
- Violence and gore
- Prohibited objects and symbols
- OCR-based text-in-image checks
- Logo and watermark detection
- Custom label training (Enterprise)
Built for scale and trust
Designed for platforms that need fast, explainable decisions.
Real-time response < 200ms
Moderate content in the request path without adding perceptible latency.
Multi-language support
Native classification across English, Chinese, Japanese, Malay and more.
Custom rules and dictionaries
Bring your own keyword lists, categories and decision thresholds.
Detailed audit reports
Per-request scores, categories and explanations you can show your review team.
Continuously improving
Models are retrained on evolving abuse patterns, so accuracy keeps climbing.
Private and compliant
GDPR-aligned data handling with regional deployment options on Enterprise.
One endpoint, two content types
Send a simple JSON payload and get a structured verdict back.
# Moderate a text payload
curl -X POST https://api.captcha.la/v1/moderation \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"type": "text",
"content": "Hello, world!"
}'
# Sample response
{
"verdict": "pass",
"score": 0.02,
"categories": {
"spam": 0.01,
"abuse": 0.00,
"adult": 0.01
},
"latency_ms": 87
}Ship trust and safety, faster.
Start free with 10,000 moderation calls per month.