Powered by MediaExtractor from the nx-net Rust crate — extract, strip, and convert media streams from YouTube, Vimeo, direct links, Tor hidden services, and generic web pages. Memory-only streaming via AudioVideoStream with zero forensic footprint.
MediaExtractor.enforce_no_local_storage() guarantees all media lives exclusively in process memory. Never written to disk. Auto-purge after configurable TTL.
MediaExtractor automatically detects .onion URLs and routes through SOCKS5 Tor proxy via nx-net. Per-request circuit isolation.
MediaExtractor.classify_source() handles YouTube, Vimeo, direct links, generic pages, and .onion hidden services. 8 container formats supported.
MediaExtractor.stream_statistics() provides real-time counters for active, completed, and failed streams. Total bytes buffered and extraction history.
Powered by MediaExtractor via extract_from_url(url) in the nx-net crate. Discovers, parses, and streams media from any URL. All buffered data lives in process memory via enforce_no_local_storage() — no disk I/O is performed. Supports YouTube URL parsing (watch, youtu.be, embed, shorts, live), Vimeo, direct links, and .onion hidden services.
MediaStream from the nx-net crate buffers bytes in a Vec<u8> that lives exclusively in process memory. MediaExtractor.enforce_no_local_storage() will panic if a disk path is ever referenced. Live progress tracking per stream.
| Stream | Source | Format | Quality | Progress | Status |
|---|
Select preferred container format and quality tier. MediaExtractor will automatically select the best available match from the source based on ExtractionConfig settings in the nx-net crate.
MediaExtractor via classify_source(url) in the nx-net crate automatically detects the source type from the URL and applies the correct extraction strategy. Tor routing is automatically enabled for .onion addresses.
Full playback with sound via AudioVideoStream created by MediaExtractor.create_av_stream() in the nx-net crate. Multi-track support for video, audio, and subtitle StreamTrack entries. All data lives exclusively in process memory — stream_only is enforced at runtime and panics on violation.
MediaExtractor in the nx-net crate classifies all discovered media into 6 distinct MediaKind variants. Each kind has specialised handling for extraction, streaming, and metadata processing.
Fine-tune extraction behaviour via ExtractionConfig in the nx-net crate. All settings apply immediately to the next MediaExtractor.extract_from_url() call.
Complete log of all MediaExtractor operations. History is stored in memory only via the nx-net enforce_no_local_storage() guarantee and cleared when the session ends — no persistent records.
No extractions yet. Use the extraction panel to get started.