Can ChatGPT watch a YouTube video?
It works with text. To work with everything said in a video, give it a transcript of the video.
Can ChatGPT transcribe a YouTube video?
Not the audio of a long video. ChatGPT works with text, so give it a transcript: paste the YouTube link here, then bring the text back to ChatGPT.
- Chat assistants work from captions or text you provide
- WhoScribe turns the audio into text, with or without captions
- Videos up to 10 hours, returned as one transcript
- Copy or export the text and paste it into ChatGPT
- $5 for a whole video, first part free
Two steps: transcript first, then ChatGPT
When you give ChatGPT a YouTube link and ask for a transcript, the answer depends on the video. If captions exist it may be able to work with them; for a long video without captions it will usually ask you to upload the audio or paste a transcript, because it does not transcribe the audio of a long video from a link. The reliable way is two steps. First, paste the link into WhoScribe: the audio is transcribed, captions are not needed, and the whole video comes back as one transcript with timestamps. Second, copy the text or download it as TXT and paste it into ChatGPT with your question, for example a summary, a blog post or a list of quotes. Each transcript on WhoScribe also has its own chat, so you can ask questions about the video without leaving the page.
How it works
Three steps to your first transcript
-
Drop in your file
Audio or video of any length; the free plan transcribes the first 15 minutes of each file. MP3, MP4, M4A, WAV, MOV and a dozen more.
-
Pick the language
Leave the language on automatic or set it for better accuracy. Turn on speaker recognition if you need to know who said what.
-
Read, edit, export
The transcript opens with clickable timestamps — read it and fix a name straight away. A free account keeps it and exports it as TXT; DOCX, PDF, SRT and VTT come with the unlimited plan.
FAQ
Questions before you upload
Try it on this pageWhy does ChatGPT ask me for the audio file?
Without captions there is no text to read from the link, so the audio has to be transcribed first.
How do I get the transcript?
Paste the YouTube link into WhoScribe. The first 5 minutes are readable with no account and the first 15 with a free account; $5 unlocks the whole video.
What about a 3-hour video?
Up to 10 hours per video works, returned as one transcript. A very long transcript may be too long to paste into a chat at once; paste it in parts or use the chat on the transcript page.
Do I need captions?
No. WhoScribe transcribes the audio, so videos without captions work.
Price, limits and formats
| YouTube and other links | Paste a YouTube link: the audio itself is transcribed, so videos without captions work, up to 10 hours per video. Links from TikTok, Instagram, Facebook, SoundCloud, Google Drive work too. |
|---|---|
| One file, no subscription | $5 unlocks the whole file, whatever its length up to 10 hours. One payment, nothing renews. |
| Try before paying | No account: the first 5 minutes of each file, 1 file a day. Free account: the first 15 minutes of each file, 3 files a day, 90 minutes a month. |
| Unlimited plan | $15 a month or $90 a year: files up to 10 hours and 10 GB, 50 files at once. (campaign price until October 31) |
| Export formats | Unlimited plan and any unlocked file: TXT, SRT, VTT, DOCX and PDF. Free plan: TXT. |
Start with one file
No account, no card. See the transcript before you decide anything.