Platform guide
How to download Facebook videos with yt-dlp
yt-dlp downloads public Facebook videos, page videos, and Reels from the normal page URL. The two things that trip people up are login-walled posts (which need cookies) and Facebook changing its page markup, which shows up as a "Cannot parse data" error until the extractor catches up.
Quick answer
yt-dlp "https://www.facebook.com/watch/?v=10153231379946729"Pass the URL of the page where the video plays. All three common forms work:
facebook.com/watch/?v=<id>— Watch videosfacebook.com/<page>/videos/<id>/— page and profile videosfacebook.com/reel/<id>— Reels (handled by the dedicated reel extractor)
Pick the quality: Facebook exposes sd and hd
Unlike YouTube's long format list, Facebook usually serves just two: sd and hd. List them, then ask for the one you want by name:
yt-dlp -F "https://www.facebook.com/watch/?v=10153231379946729"
# ID EXT RESOLUTION | PROTO | VCODEC ACODEC
# sd mp4 unknown | https | ...
# hd mp4 unknown | https | ...
yt-dlp -f hd "https://www.facebook.com/watch/?v=10153231379946729"yt-dlp already picks hd by default when it is available, so -f hd is only needed if you want to force it or fall back to sd on a slow connection.
Private, group, and login-walled videos need cookies
Public videos download anonymously. Anything behind a login — private profiles, group posts, some regional content — only extracts if you pass the session from a browser where you are logged in:
yt-dlp --cookies-from-browser firefox "https://www.facebook.com/<page>/videos/<id>/"See the cookies guide for exporting from Chrome or a cookies file instead.
Fixing "Cannot parse data" and sudden breakage
Facebook changes its page structure often, and when it does the extractor can return ERROR: Cannot parse dataor a wrong "login required" message on public videos (see yt-dlp issues #17906 and #18016, both filed in late 2026). The fix order is almost always the same:
- Update first. A stale build is the most common cause. Update to the latest release — see updating yt-dlp. Nightly sometimes has the Facebook fix before the stable release does.
- Retry with cookies. Some "Cannot parse data" cases are really a soft login wall; add
--cookies-from-browser firefox. - If an up-to-date build still fails on one specific video, it is extractor lag, not your setup. Capture verbose output with
-vand check or file it on the error-triage guide.
If it keeps breaking
Tired of fixing this every time a site changes?
Paste a link to see what a managed import API returns — the one that stays current so your workflow doesn't break.
Running it on a server?
Facebook extraction that works on your laptop can fail from a cloud box: datacenter IPs draw more checks, and the cookies you exported locally are not on the server. If you are hitting this in production, the yt-dlp blocked on a server guide covers the options.