Skip to main content

Incorrect / Mismatched Transcript Returned for YouTube Video

Feature Request

We are facing an issue with the YouTube transcript generation for a specific video ID using your transcription service.

Issue Description: For the provided YouTube video ID, the transcript returned by your system appears to be incorrect and does not correspond to the actual video content.

Observed Behavior:

  • The YouTube video duration is approximately 2 hours.

  • However, the transcript returned is only around 25 seconds long.

  • The transcript content does not match any part of the original video.

  • The returned transcript appears to be unrelated or possibly from a different video (fake/incorrect transcript).

Posted by Vipul Citrusbug ·
Status: Completed7 comments

Log in to comment and vote

Comments7

  • info

    •

    Mar 26

    •

    Merged request

    •

    1 vote

    Getting the Wrong Transcript

    We need your support on a transcript accuracy issue.

    For YouTube video ID byk3W3Xpn4k, the transcript returned by your API is incorrect.
    We manually checked the video and the returned transcript does not match the spoken content.

    Example of returned text:
    “I’m a computer scientist ... using artificial intelligence to make things more accessible ... tools for blind or low-vision people...”

    Two observed retrieval attempts (UTC):

    2026-03-12 13:57:40+00

    2026-03-12 14:02:39+00

    Integration details:

    Platform: Node.js backend service

    Integration type: direct HTTPS REST call (no SDK)

    Endpoint pattern: GET {SUPADATA_BASE_URL}/transcript?url=https://www.youtube.com/watch?v=byk3W3Xpn4k&lang=en&text=false&mode=native

    Auth header: x-api-key

    Could you please investigate:

    Why transcript content for byk3W3Xpn4k does not match actual video audio.

    Whether there were any source-mapping/retrieval issues around those timestamps.

    and, what’s more concerning is : I am not receiving any reply from support via the Chat → Send Us a message. The bot was saying a teammate will be responding. I left this message on 13rd March and today is 23rd March already. Not even a courtesy reply ?

  • james

    •

    Mar 17

    I hit this with 3 videos today: 3h25PrfZFL4, 3rSs8kLlppo, vtBIulnRrSA

    These are all long public meetings (eg city council) that were live-streamed last night.

    If I don't process these vids for a day, does anyone know if that gets things straightened out?

  • info

    •

    Mar 12

    I ran into a similar issue today. I was processing a finance related video that is 47 minutes long but the transcript returned was a totally irrelevant paragraph :

    “The solar system is a vast and mysterious place, consisting of a central star, the Sun, and all the objects that are gravity bound to it. These include the eight major planets, their moons, dwarf planets, and countless smaller bodies like asteroids and comets. At its heart is the Sun, a massive ball of glowing gas that provides the energy necessary for life on Earth. Exploring these celestial bodies helps us understand our origins and the nature of the universe.”

    you guys eventually got any feedback on how to fix the problem?

  • ethan

    •

    Feb 28

    This is also a blocking issue for us. We’re seeing it on any live stream that hasn’t yet started. Expected behavior is that we should get an error or an empty transcript. Instead we get transcripts for completely unrelated videos. Once the video is available, it appears that maybe you’re already caching the incorrect transcript, and continue to serve it.

    Examples:

    https://www.youtube.com/watch?v=yG3ykQ70MIo (Aired on Feb. 26, but we probably had already tried to transcribe it before it was live, and now the incorrect transcript is still returned)

    https://www.youtube.com/watch?v=96RDF14r2ws (Will be live on March 5)

    https://www.youtube.com/watch?v=40xLm9nbwKU (Will be live on March 4)

  • humberto.soares

    •

    Feb 19

    •

    Merged request

    •

    2 votes

    Incoerent transcription from some Youtube videos

    https://www.youtube.com/watch?v=HY_JyxAZsiE
    If I try to get this video transcription, it come with a small text, that it seems like a summary of the video… Do you have any caches for the trnascripts? It is only for some videos…

  • tls.investments.llc

    •

    Feb 19

    •

    Merged request

    •

    2 votes

    Incorrect Transcript Scraping

    "Your API at https://api.supadata.ai/v1/youtube/transcript returns the wrong transcript for video Veu2T-q7Yac. The video is about Bitcoin, but your API returns a 26-word JSON schema tutorial instead of the actual 11,000+ word transcript."

    https://www.youtube.com/watch?v=Veu2T-q7Yac

  • Vipul Citrusbug

    •

    Feb 5

    Here is the video URL :

    https://www.youtube.com/watch?v=qUeIUtj-gmg