1. Paste a public HTTPS link
The link is normalized, stripped of common tracking parameters, resolved against approved platform adapters, and checked against private and internal network ranges.
2. Choose an honest format
The extraction worker reads metadata without downloading the complete media. Malformed, DRM-protected, live-only and duplicate formats are excluded. The UI shows only qualities the source actually reports.
3. Prepare and download
A background job downloads the selected stream, combines audio when needed, validates the output with FFprobe and uploads it to private temporary storage. Your browser receives a short-lived download redirect, never storage credentials.
Automatic cleanup
The file is scheduled for deletion when it becomes ready. Retry and reconciliation jobs handle missed deletion attempts, and local worker directories are always removed.