SERA.toolkit
A universal media toolkit for downloading and converting publicly accessible media.
How it works
Paste a link. SERA works out which site it belongs to, asks that site what media the post contains, and offers you the formats that genuinely exist for it. There is no source picker, because there is nothing useful to pick from until the link has been read.
A post with several images or videos resolves to all of them, not just the first. Pick the parts you want; several files come back as one archive, or separately if you would rather.
Supported sources
The list of supported sources is read from the API, which is not reachable right now.
Formats
Video comes back as MP4, WebM or MOV — whichever the source can be put into without re-encoding — at whatever resolutions the site actually publishes. Audio can be extracted as MP3, M4A, Opus or WAV. Images come back untouched, and short videos can be converted to GIF.
A quality is only offered when the source has it. If the original audio is 128 kbps, SERA will not offer to hand you a 320 kbps file, because there would be nothing extra in it.
Privacy
There is no account and no tracking. Nothing you paste is stored: the link is used to fetch the media and then discarded, and the server logs record which site a request went to and whether it worked, not which video you asked for.
Finished files live in a temporary directory and are deleted automatically, whether or not you downloaded them. Thumbnails are fetched by the server rather than your browser, so the site you pasted from never sees your address.
Responsible use
You are responsible for making sure you have the right to download and convert whatever you put through SERA. Copyright, platform terms and local law all still apply, and nothing here changes them.
SERA only handles media that is already publicly accessible. It does not defeat DRM, sign in on your behalf, bypass paywalls, or reach content behind an access control — if a post is private, SERA will tell you so rather than trying to get around it.
Open source
SERA.toolkit is MIT licensed and self-hostable. It is built on yt-dlp for extraction and FFmpeg for conversion, both of which do the genuinely hard work here.
Run it yourself with Docker, or from source. The README covers both.