Image Reader

OCR-based text extraction from images, applying your moderation rules to the extracted content.

How It Works

Image Reader uses Optical Character Recognition (OCR) to extract visible text from images and screenshots. The extracted text is then run through your configured Auto Moderation rules (toxicity, spam, links, phishing, etc.).

Use Cases

Screenshot Toxicity

Catch users posting screenshots of toxic messages or slurs to bypass text filters.

Phishing Images

Detect phishing URLs embedded in images or fake login page screenshots.

Advertisement Images

Catch promotional content and invite links embedded in images or banners.

Rule Evasion

Prevent users from converting blocked text into images to evade moderation.

OCR processing adds a brief delay (typically under 2 seconds). Images are not stored — extracted text is discarded after evaluation.