Get a YouTube transcript when the video has no captions
If YouTube shows no “Show transcript” option, the video has no caption track, and tools that copy captions return nothing. Paste the link here and the audio itself is transcribed, so the missing captions do not matter.
Works when “Show transcript” is missing
Transcribes the audio, not the caption file
Videos up to 10 hours
Speaker labels and timestamps
First 15 minutes free, whole video for one payment
Start a transcript
Paste a link to a video or recording, or upload the file.
Have the file instead?
Upload a fileDrop it to add itDrop audio or video here · MP3, MP4, M4A, WAV, MOV and more
·
OptionsLanguage, speed, English copy
No matching language.
Picking the language improves accuracy. Leave it automatic if you are unsure.
Model quality
Why a YouTube video has no transcript, and what to do
YouTube’s transcript panel only shows captions that already exist. A video has none when the creator has turned captions off, or when YouTube’s automatic captions were never generated. YouTube’s own help page lists the usual reasons: the language is not supported by automatic captions, the video is too long, the sound quality is poor, the video starts with a long silence, or several people speak at once. Livestream recordings and videos uploaded in the last few hours often have no captions yet either. In all of these cases a caption extractor has nothing to extract, which is why it answers “transcript not available”. The fix is to transcribe the audio instead of looking for a caption file. That is what WhoScribe does with a YouTube link: it reads the audio of the public video and produces the text with speaker labels and timestamps. This is not a rare case in our own data. Of the last 69 YouTube videos that people paid WhoScribe to transcribe in full, 34 had no captions of any kind when we checked on 4 October 2026, and their median length was 68 minutes.
Before anything else
First check: is there really no transcript?
Thirty seconds on YouTube can save you the whole job. On a computer:
1
Open the description
Click “…more” under the video to expand the full description. Some creators paste the transcript there, or link to one.
2
Look for “Show transcript”
At the end of the expanded description, a “Show transcript” button appears when the video has captions. It opens the text beside the video, with timestamps (YouTube Help: View video transcripts).
3
Try another caption language
Open the player’s Settings (the gear) and then Subtitles/CC. If any track is listed, even an auto-generated one or one in another language, the video does have captions, and the transcript can be shown in that language.
If none of the three turns anything up, the video has no caption track, so copying captions cannot work. YouTube’s help page on automatic captions lists the usual reasons. The way out is to transcribe the audio itself: on your own computer (below) or here, without installing anything.
The free route
Do it yourself, free: yt-dlp + Whisper
If you are comfortable in a terminal, two free, open-source tools do the whole job on your own computer: yt-dlp downloads the audio of the video, and OpenAI’s Whisper turns that audio into text. Nothing is uploaded anywhere.
What you need, honestly
Python. Whisper’s README says it is expected to work with Python 3.8–3.11, and yt-dlp’s README supports 3.10 and newer, so Python 3.10 or 3.11 suits both.
ffmpeg (the program, not the Python package of the same name). Both tools need it; install it with your package manager, for example brew install ffmpeg on a Mac or sudo apt install ffmpeg on Ubuntu.
For YouTube, yt-dlp’s README also recommends a JavaScript runtime such as Deno. Installing yt-dlp with its “default” extra adds the matching component.
Disk space: Python packages, the Whisper model (downloaded the first time you use it) and the audio file. Plan for a few GB.
A graphics card (GPU), or patience. The table below shows the video memory (VRAM) each model needs. Without a GPU, Whisper runs on the processor, so a long video needs patience.
Replace <URL> with the video link. -x keeps only the audio, and -o names the file audio.mp3 (without it, yt-dlp names the file after the video title). Only download videos you have the right to download.
Whisper writes audio.txt next to the audio file. Use --output_format srt for subtitles with timestamps, and add --language Spanish (or another language) if Whisper guesses the language wrong.
Whisper models and the video memory they need, quoted from the openai/whisper README (checked )
Model
Parameters
Name for --model
Required VRAM
Relative speed (vs large)
tiny
39 M
tiny
~1 GB
~10x
base
74 M
base
~1 GB
~7x
small
244 M
small
~2 GB
~4x
medium
769 M
medium
~5 GB
~2x
large
1550 M
large
~10 GB
1x
turbo
809 M
turbo
~6 GB
~8x
The README measured the relative speeds on English speech on an A100 GPU and notes that real-world speed varies with the language, the speaking speed and the hardware. It describes turbo as an optimised version of large-v3, faster with a minimal loss of accuracy, but not trained for translation.
What you get: a free, private transcript. What it costs: the first setup, your computer’s time on long videos, and a plain text file with no editor.
The no-install route
Or paste the link here: no install
Paste the YouTube link into the box at the top of this page. WhoScribe fetches the audio and transcribes it on its own servers, so there is nothing to install and no GPU to own.
Videos up to 10 hours long
The first 5 minutes are readable with no account, and the first 15 minutes with a free account
$5 unlocks the whole video: one payment, no subscription
Public videos only (anyone must be able to open the link without signing in)
Transcribe a YouTube video
Paste the link. The first 5 minutes are free without an account, 15 with a free account. The whole file is $5, a one-time payment.
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.
2
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.
3
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.
Why is there no “Show transcript” button on my video?
Because the video has no caption track. Either the creator disabled captions or YouTube did not generate automatic ones. YouTube lists unsupported language, a very long video, poor audio, a long silence at the start and overlapping speakers as the common causes.
Can I get a transcript anyway?
Yes, by transcribing the audio. Paste the video link into WhoScribe; it does not look for captions, it listens to the audio and writes the text. The video has to be public.
Will caption-based transcript tools work on a video like this?
No. Tools that copy YouTube’s captions depend on a caption track existing. When there is none they return an error or an empty result. Only speech-to-text on the audio works.
How long can the video be?
Up to 10 hours per video. Long lectures, podcasts, sermons, hearings and audiobooks are common here.
What does it cost?
The first 5 minutes are free with no account and the first 15 minutes with a free account. One payment of $5 unlocks the whole video ($3 in most lower-income countries), with no subscription. Prices as of 4 October 2026.
Which languages work?
Over 90 spoken languages, including many that YouTube’s automatic captions do not cover. The finished transcript can also be translated.
Does it work on private, members-only or age-restricted videos?
No. Only videos that anyone can open without signing in. For a video you own, download the file from YouTube Studio and upload it instead.
Can I transcribe a livestream?
After it ends and the replay is published, yes. A stream that is still live cannot be imported.
Working in another language?
All of this works in 100 languages. These are the ones asked for most, each with its own page.