Content Moderation
Flag potentially unsafe or inappropriate images before they reach your users or moderators. Each image is screened for sensitive categories and returns the ones it finds, with a confidence score, so you can auto-block, blur or route for human review.
What Content Moderation does
Detects categories such as explicit and suggestive nudity, violence, visually disturbing content, drugs, tobacco, alcohol, gambling, hate symbols and rude gestures
Hierarchical labels: a top-level category plus a specific sub-label, each with its own confidence score
Set a minimum confidence to tune how sensitive the filter is
Use as an automated pre-filter, then send only borderline cases to human reviewers
Privacy-first, the image is processed for the check and then discarded
Frequently asked
What content can it flag?
Sensitive categories including explicit and suggestive nudity, violence, visually disturbing imagery, drugs, tobacco, alcohol, gambling, hate symbols and rude gestures. It analyses image content only, it does not judge intent or context.
Is it accurate enough to block automatically?
It returns a confidence score per label, so you choose a threshold. The common pattern is to auto-handle high-confidence results and route borderline ones to a human reviewer, rather than relying on it alone.
Can we tune how strict it is?
Yes. Set a minimum confidence: a higher bar means fewer false positives but may miss borderline content, a lower bar catches more but flags more for review.
Is any image stored?
No. By default the image is processed for the moderation result and then discarded.
Add Content Moderation to your flow
One API request. Try it live in your browser, or get in touch.
eaglecheck, one API for age, identity and image checks.