Paste a Facebook post, Reel or Live replay link and get the spoken words as text, free when the video already has captions.
Get a Facebook transcript →Facebook can display captions, but whether a given post has them is up to whoever uploaded it. Some pages attach a caption file. Some switch on the automatic ones. Plenty do neither.
So a transcript tool needs two paths. JotVox looks for a caption track first and pulls it as-is. With nothing to pull, it transcribes the spoken audio instead.
Where captions do exist, they are not always the version you want. Automatic text struggles with accents, crosstalk and speech under music, and it arrives lightly punctuated with no speaker labels.
Because feed video is mostly watched muted, creators often burn text into the frame instead of attaching a track. Burned-in text is part of the picture, so those posts count as having no captions.
Pages behave differently from personal profiles, too. A publisher running video at scale usually has a captioning step in its workflow; someone filming a reply in a group does not. That is why two Facebook links can behave completely differently.
Reels are the format people ask about most, and the fiddliest: short, loud, and usually sitting on a music bed.
Music does not block a transcript; someone talking over a beat comes out fine. But a Reel that is thirty seconds of trending audio with on-screen text has nothing to transcribe.
The rest of what makes Reels awkward is pace. People speak fast to fit a hook, a point and a call to action into under a minute, and cut hard between sentences, so words get clipped at the join. A voiceover recorded into an editor comes out cleaner than phone audio shot at arm's length in the street.
Open the Reel, tap share, copy the link, paste it. Most fb Reels carry no caption track, so they run on AI transcription. The same clip usually lives on Instagram and TikTok, and those links work identically.
What people do with a Reel transcript is narrower than with longer video: read the hook back to work out why it landed, translate the script for another market, or build a subtitle file for a re-upload. At thirty seconds it is short enough to tidy by hand.
Live is the opposite problem: length. A service, a council meeting or a two-hour community Q&A sits on the page afterwards as a replay, and nobody wants to scrub it for the one announcement that mattered.
Once the broadcast has ended and the replay is posted, paste its link like any other video. You get the full spoken record, with timestamps if you export SRT or VTT.
A stream still running is different: no finished recording exists yet, so wait for it to end. For long fb live replays, an AI summary earns its credit, turning two hours into chapters and takeaways.
Live audio has its own texture. Broadcasts open with several minutes of waiting for people to arrive, the host reads comments aloud, and guests dial in with worse audio than the room. All of it lands in the transcript in order, so expect to delete the first stretch.
For a long replay, export SRT rather than TXT. Timestamps let you find the two minutes where the date, the price or the decision was announced, quote it exactly, and point people at that moment rather than the whole broadcast.
Plenty of people searching for a Facebook transcript are not archiving anything. They want the script, the actual spoken words, because a video performed and they want to rebuild it, translate it, or write the next one on the same bones.
TXT is the export for that: no timestamps, no cue numbers, just the words in order, ready to paste into a doc and edit.
Be clear about what that gives you: the words as spoken, not a shooting script. Filler, restarts and asides are in there, and nothing on screen is. Cleaning it up takes one editing pass, which still beats typing it out by ear.
For the shape rather than every word, an AI summary returns takeaways, quotes and a chapter breakdown.
Community managers lift a Live announcement into meeting notes so people who missed it get the detail, not a paraphrase. Creators pull the script from a Reel that worked and rebuild it elsewhere. Marketers transcribe a rival's video ad to see how the offer is worded.
Groups are their own case. A long explainer in a hobby group is often the only place that information exists, and it only ever gets spoken. A transcript makes it searchable. Same story on LinkedIn.
Whole categories of page live on spoken video. Churches post the service. Local government pages post the meeting. Small businesses answer customer questions on camera instead of writing it down. The only text version is the one someone makes afterwards.
Accessibility is the other reason. Followers who are deaf or hard of hearing get nothing from an uncaptioned post, and an SRT alongside a re-upload closes that gap.
Not every URL containing a video points at one. The link you want opens the post itself: from the post's own timestamp, the three-dot menu, the share icon on a Reel, or a page's list of videos.
One test settles the rest. Open the link in a private browser window. If the video plays for a signed-out visitor, it will paste fine. If you hit a login prompt, uploading the file is the answer instead.
Embedded players are the other case. A Facebook video embedded in a news article is still hosted on Facebook, so click through to the post and copy that link, not the article's.
Grabbing captions that already exist is free and needs no sign-up. New accounts also get 3 free credits plus one free AI transcription a day after email verification. Packs start at $4.99 for 20 credits.
One credit is one action, so a two-hour Live replay costs the same as a thirty-second Reel.
Two Facebook-specific traps are worth naming. A link copied out of the mobile app sometimes resolves to a login wall rather than the post, so test it signed out first. And a crossposted video may have a cleaner original elsewhere, which is often the better link.
1. Copy the post link Use the three-dot menu on a post, the share icon on a Reel, or open a finished Live replay and copy its link.
2. Paste it into JotVox No account is needed to check whether the post already carries captions you can take for free.
3. Take the captions, or transcribe A caption track comes back free. If there is none, AI transcription reads the audio in 99 languages.
4. Export or summarise TXT for a clean script, SRT or VTT for timestamps, MP3 for audio, or an AI summary for takeaways and chapters.
Paste the post link into JotVox. If the video already has a caption track, pulling it costs nothing and needs no account. Only AI transcription uses credits, and new accounts get 3 free ones plus a free AI transcription every day after email verification.
Yes. Open the Reel, use the share icon to copy its link, and paste it in. Most Reels have no caption file, so they run through AI transcription, which copes with a music bed under a voice. A Reel that is only trending audio and on-screen text returns nothing.
Yes, once the broadcast has ended and the replay is on the page. Copy the replay link and paste it. A stream still running has no finished recording, so wait. Long fb live replays are where an AI summary pays for itself, turning hours into chapters and takeaways.
No. A private, friends-only or closed-group post cannot be reached without logging in as someone with access, and JotVox does not sign into accounts. If you have access and the right to use the video, download it and upload the file from your device, Drive or Dropbox.
Export the transcript as TXT. That gives you the spoken words as plain text with no timestamps or cue numbers, which is what most people mean by a script download. Paste it into a doc and rewrite the hook, translate it, or reuse the structure.
Not quite. Captions are timed text shown over the video as it plays. A transcript is the full text on its own, in reading order. JotVox gives you both: SRT or VTT for timed caption files, TXT for a flat transcript, from the same paste.
AI transcription covers 99 languages, which matters on Facebook because so much page and group video is not in English. A Portuguese Reel or a Spanish Live replay is transcribed in the language spoken, at 98% accuracy or better on clear audio.