Facebook Video Transcript API: Extract Reel and Post Text

A Facebook video transcript API converts the speech in a public Facebook video or Reel into plain text. The result can be searched, summarized, translated, indexed, or passed to an LLM without downloading and transcribing the media yourself.

CreatorCrawl exposes this as GET /api/facebook/post/transcript.

Get a Facebook video transcript

Pass the public post or Reel URL and your API key:

curl "https://app.creatorcrawl.com/api/facebook/post/transcript?url=https%3A%2F%2Fwww.facebook.com%2Freel%2FVIDEO_ID" \
  -H "x-api-key: YOUR_API_KEY"

A successful response places the upstream transcript fields inside the standard envelope:

{
  "data": {
    "transcript": "First line of the video...\nSecond line..."
  },
  "meta": {
    "platform": "facebook",
    "fetched_at": "2026-08-31T10:00:00.000Z"
  }
}

The current upstream endpoint supports public videos under two minutes. Private, deleted, gated, or non-video posts cannot return a transcript.

Python example

import requests

response = requests.get(
    "https://app.creatorcrawl.com/api/facebook/post/transcript",
    headers={"x-api-key": "YOUR_API_KEY"},
    params={"url": "https://www.facebook.com/reel/VIDEO_ID"},
    timeout=60,
)
response.raise_for_status()

transcript = response.json()["data"].get("transcript")
if transcript:
    print(transcript)

Check the video before transcribing

Call the post endpoint first when you also need engagement metrics, author information, the video URL, thumbnail, duration, or a faster comment lookup identifier.

curl "https://app.creatorcrawl.com/api/facebook/post?url=FACEBOOK_VIDEO_URL" \
  -H "x-api-key: YOUR_API_KEY"

The post response can include:

• Post ID and canonical URL
• Description and creation time
• Like, comment, share, and view counts
• SD and HD video URLs
• Thumbnail and duration
• Author details
• feedback_id for comment retrieval

That makes a two-request workflow useful for building a video corpus: fetch metadata, then request the transcript only for qualifying videos.

What to build with Facebook transcripts

Creative research

Collect transcripts from public brand and creator Reels, then group hooks, offers, objections, and calls to action. Text is easier to compare than raw video.

Searchable video archives

Store the post URL, author, publish date, engagement metrics, and transcript in your database. Add full-text or vector search so a team can find every video that mentions a product, topic, or claim.

Caption and accessibility workflows

Use the transcript as a draft for captions, summaries, translations, or internal notes. Human review is still appropriate when wording matters.

AI-agent research

Pass transcript text to an LLM with the source URL and metadata. Ask for a summary, claims, named products, sentiment, or recurring themes across a group of videos.

Transcript API versus downloading the video

Downloading media and running speech recognition gives you more control, but it also requires file handling, audio extraction, model hosting or transcription fees, retries, and cleanup. A transcript endpoint is the smaller integration when text is the only output you need.

For the rest of the Facebook data surface, see the Facebook scraper API guide and Facebook API hub. Create an API key to test the transcript endpoint with free credits.