Turn a talk into a script
Transcribe a webinar or a YouTube video, tidy the text, and use it as the script for a new narration in our text to speech voices.
Transcripts from video
Speech in, editable text out
Whisper Large V3 Turbo listens to the audio track and returns every spoken line with its start and end time. You check it against the video, fix what it misheard, and keep the result as plain text or as caption files.
Related use cases
The quick answer
Open Video to Text on Auto Caption, drop your MP4, MOV or WebM, wait for the lines to appear, fix any wrong words, then download TXT for a transcript or SRT and VTT for captions.
4 steps
How to transcribe a video to text
- 01
Open the video
Drop an MP4, MOV or WebM up to 30 minutes. Your browser opens it on your device; the video file itself is never uploaded.
- 02
Let Whisper listen
The browser pulls out a small, compressed copy of the speech and sends only that to Whisper Large V3 Turbo. Most clips come back in under a minute.
- 03
Read and fix
Every line is timed. Click one to jump the video there, listen again, and correct a name or a word the model misheard.
- 04
Download the text
Take a plain TXT transcript, or SRT and VTT files with timestamps for captions. Downloading again after edits is free.
Why transcribe
What people do with a video transcript
Quotes and show notes
Find the exact sentence someone said without scrubbing through the timeline. Search the transcript, click, and the video jumps there.
Captions for a finished cut
The same timed lines export as SRT or VTT for YouTube, Vimeo or LinkedIn. Need them on the picture? Style them and render an MP4.
Repurpose screen recordings
Product demos and tutorials become help articles faster when the spoken steps are already typed out.
Video to text runs on Auto Caption, our free caption and transcription site. It uses the same Whisper engine as our Audio to Text tool, with an editor built for video.
Free video transcription, with the limits up front
Without an account you can transcribe 20 minutes of video a day, and 60 minutes once you sign in. A single file can be up to 30 minutes long. A 12-minute video uses 12 minutes; the allowance refills as each use turns 24 hours old. Everything after the transcription is free and unlimited: reading, correcting, searching and downloading the text again.
The speech model is Whisper Large V3 Turbo running on Groq, the same engine behind our Audio to Text tool. Clear speech close to the microphone comes back nearly clean. Music under the voice, two people talking at once, and unusual names cause most mistakes, which is why the transcript opens in an editor rather than a text box you can only copy.
MP4 to text, and the formats you get back
MP4 to text is the common case, but MOV from an iPhone and WebM from a screen recorder work the same way. When the words are ready, the Export panel offers four files:
- TXT: the words only, one caption line per row. Best for notes, articles and scripts.
- SRT: numbered, timed blocks. YouTube, Facebook, LinkedIn and most video editors read it.
- VTT: the web's caption format, for video players on your own site.
- SBV: YouTube's own caption format, for older upload workflows.
Before you publish captions, run the SRT through Caption QA to catch lines that flash by too fast or overlap.
Transcribe a video, then give it a new voice
A transcript is often the first step of a re-voice: a talk recorded in a noisy room, a tutorial with a script that changed, or a clip you want in a clearer voice. Fix the transcript on Auto Caption, paste it into free text to speech, and download the narration. If you kept the SRT, SRT to Speech generates narration that follows the original timing, and Voiceover for Video lets you place it back over the picture.
Free on Auto Caption
Ready to try it on your own video?
Open the free editor on Auto Caption, drop your file, and download the result in a few minutes.
Transcribe a video
Opens auto-caption.app
FAQ
Video to Text FAQs
01 Is the video uploaded to your server?
No. The video is opened by your browser. To transcribe it, the browser extracts a compressed mono audio track and sends only that to the speech model. We keep a usage record (length and time) for the daily allowance, not the audio or the transcript.
02 How much video can I transcribe for free?
20 minutes a day without an account and 60 minutes a day when you sign in, with files up to 30 minutes long. Minutes come back as each use turns 24 hours old. Editing and downloading the text again never uses minutes.
03 Which video formats work?
MP4, MOV and WebM, plus MKV and M4V in most browsers. If a file will not open, re-save it as MP4 (H.264 video, AAC audio), which every browser reads.
04 Which languages are supported?
Whisper recognises 50+ languages and detects the spoken language automatically. Pick the language yourself if the clip is short, starts with music, or mixes languages.
05 Does it label speakers?
No. You get the words in order with timestamps. For an interview, add names while you read through; every line can be edited.
06 I only have an audio file. Can I still use this?
Yes, drop an MP3, M4A or WAV into Auto Caption and it works the same way. Our Audio to Text tool is also built for recordings and exports the same formats.
Still wondering? Get in touch →
Related tools
Keep going
Recordings
↗
Audio to Text
Transcribe MP3, WAV and M4A recordings into editable text and captions.
Captions
↗
Caption QA
Check an SRT or VTT for overlaps, fast lines and long rows before you publish.
Voice
↗
SRT to Speech
Turn a subtitle file into timed narration in 54 natural voices.
Voice
↗
Voiceover for Video
Add new narration to your video and preview the timing before you export.
Try it now
Turn the transcript into a voiceover
Paste your edited script into the text to speech tool, pick one of 54 natural voices, and download MP3 or WAV.
