Home / Written Caption vs Transcript

Can a Reel Transcript Tool Copy the Written Post Caption?

No. XAPN turns the spoken audio of one public Instagram Reel into text; it does not copy the creator-written caption beneath the post. If you need the words you wrote in your own post, retrieve that written caption from the post or your own draft. Use the transcript tool when you need what was said aloud, then review its wording before reuse.

A Reel has separate written post text and spoken audio. XAPN's checked route processes audio into a transcript, not the written caption.
Two text sources, two different retrieval jobs.

What we checked on this site

On October 12, 2026 (Los Angeles time), we read the JavaScript served by XAPN's live homepage and compared it with the local site file. The page sends the pasted Reel URL to /audio. The service route uses an audio-only media choice, then returns audio bytes to the browser. There is no request for a post description in that route. This is a check of XAPN's implementation, not a claim about every transcript tool.

In the browser, the returned audio is decoded and resampled before automatic-speech-recognition runs with the English Whisper tiny model. The result area receives out.text.trim(). Copy writes that result area's text to the clipboard; Download .txt builds a file from the same text. We found no separate field, request, or merge step for the post's written caption in this path. A successful download therefore preserves the speech-recognition result, not an archive of what the creator typed below the Reel.

We also checked the live homepage and sitemap before publishing this guide: both responded with HTTP 200. The inspected live page exposed the audio request, recognition call, and result assignment. We did not log into Instagram, access a creator's account, or compare a particular post's caption with its spoken words today. A caption might happen to repeat speech, but that coincidence would not make this a caption extractor.

Evidence boundary: this is a current implementation inspection. It establishes which data XAPN requests and displays. It does not measure recognition accuracy or test whether Instagram permits copying text from every account or device.

Choose the text you actually need

  1. Your written post caption: go back to the post or the draft you control. Copy or export that source directly; transcribing its Reel audio is a different operation.
  2. Words spoken in the Reel: paste the individual public Reel link into XAPN, wait for Done. and visible words, then copy or download the result.
  3. Text drawn on the video: inspect the frames separately. This audio-based path does not read text overlays or stickers.

For a whole account's posts, see the account-link test; a profile URL is not a batch caption export. If the speech result itself will not copy, the clipboard check addresses that separate issue. Do not substitute a speech transcript for a written caption when quoting, archiving, or republishing someone's post.

Questions

Will the transcript include hashtags from the written caption? Not because they appear under the post. Only words recognized from the Reel's audio are sent to the result area.

Can I copy my own spoken script? You can copy the recognized text after it appears, but listen back and correct names or phrasing before treating it as your final script.

The written post caption and spoken audio remain different sources. You can convert a Reel to text when you need the words that were said aloud. Copy the post caption separately if that is the material you actually need.