Verify how any website is classified by our DNS filtering service. Check categories and identify misclassifications in seconds.
Enter a domain to see its current content filtering classification and parent category grouping.
Every domain passes through a multi-stage classification pipeline. Each stage contributes a confidence signal, and the final category is determined by weighted consensus across all stages.
Natural language processing examines page content, meta tags, titles, and structured data to determine the primary topic of a website.
Computer vision models analyze page screenshots, logos, and visual layouts to identify category-specific design patterns and imagery content.
Traffic patterns, DNS query volumes, linking relationships, and user interaction signals provide behavioral context for classification.
Weighted consensus across all signals produces the final category assignment. Categories are validated against our 59-category taxonomy for consistency.
Achieving 99.2% accuracy across 90 million domains requires a rigorous methodology that combines automated analysis with human oversight.
Our ensemble of transformer-based models is trained on millions of manually verified classifications. Models are retrained weekly to adapt to evolving web content patterns and emerging categories.
Domains with ambiguous classifications are routed to our classification team for manual verification by a human analyst.
The entire domain database is recrawled on a rolling 30-day cycle. High-traffic and recently changed domains are recrawled daily to ensure classifications remain current.
We maintain a gold-standard test set of 50,000 human-verified domains. Every model update is benchmarked against this set before deployment to prevent accuracy regressions.
Look up any domain against our live classification database, or compare plans to integrate category data into your own filtering stack.
Check Domain See Plans