Skip to content
Scriptorium
Transcribe free

Turn video and audio into editable, searchable, timestamped text

TryScriptorium turns audio and video into editable, timestamped text in 33 languages, including Arabic. Upload a recording, correct any line, search across everything you transcribe, and export TXT, SRT, VTT, Markdown, JSON or CSV. One file up to 15 minutes and 25 MB is free with no account.

  • No sign-up needed: transcribe a file of up to 15 minutes and download it as SRT, VTT or text, free.
  • Drop a file or record from your microphone
  • 25 MB per file, 3 tries a day, no card and nothing to install

See a finished transcript first

Live replay
00:00 / 00:24
  1. 00:00Good morning everyone, let's start with the project update.
  2. 00:04The design review is finished and the launch date is now the 12th of March.
  3. 00:10The budget is on track but we still need approval for the extra servers.
Real output: the free-preview engine (small model) on this sample, unedited.English voice: synthetic (Piper, trained on public-domain LJ Speech).

Free preview: one file up to 15 minutes and 25 MB, 3 tries a day, no account needed.

Want to keep your transcripts? Create a free account

Play it, search it, fix a word

A short recording and the transcript this product made from it. Nothing to upload.

0:00 / 0:24
Tapping a line

  1. Good morning everyone, let's start with the project update.

  2. The design review is finished and the launch date is now the 12th of March.

  3. The budget is on track but we still need approval for the extra servers.

  4. Maria will send the final numbers by Friday.

  5. Next week we will test the new export feature.

  6. Thank you all and see you on Monday.

Tap any line to play from there. Switch to Edit to fix a word.

1
00:00:00,000 --> 00:00:04,480
Good morning everyone, let's start with the project update.

2
00:00:04,480 --> 00:00:10,000
The design review is finished and the launch date is now the 12th of March.

3
00:00:10,000 --> 00:00:14,920
The budget is on track but we still need approval for the extra servers.

4
00:00:14,920 --> 00:00:18,600
Maria will send the final numbers by Friday.

5
00:00:18,600 --> 00:00:21,920
Next week we will test the new export feature.

6
00:00:21,920 --> 00:00:24,000
Thank you all and see you on Monday.

Sample: a synthetic voice reading a short script, so the clip is free to use. The transcript is the unedited output of this product's free-preview engine (small model, language detected automatically).

How it works

  1. Add a recording

    Drop in a video or audio file, choose one from your device, or record from the microphone. Nothing to install.

  2. It transcribes

    Speech becomes timestamped text, with speaker labels when more than one person talks.

  3. Read, fix, export

    Click any line to correct it, then download TXT, subtitles, Markdown, JSON or CSV.

33 languages, plus auto-detect

Pick the language, or let it work it out. Speech in any of them can also be translated into English text.

  • Arabic
  • English
  • Spanish
  • French
  • German
  • Hindi
  • Indonesian
  • Turkish
  • Urdu
  • Portuguese
  • Russian
  • Chinese
  • Japanese
  • Italian
  • and 19 more

See every language, with notes on each

Who uses it

Anywhere a recording needs to become something you can read.

  • YouTube and video creators

    Pull a transcript out of a finished cut for captions, chapters, descriptions and repurposing.

  • Podcasters

    Turn an episode into show notes and a searchable archive, with each speaker labelled.

  • Students

    Get a lecture recording into text you can search, quote and revise from.

  • Researchers

    Transcribe interviews and field recordings, then search across all of them at once.

  • Journalists

    Get an interview into quotable text quickly, with timestamps back to the moment.

  • Teams and meetings

    Turn a recorded call into a written record with who said what.

All use cases

Use it here, or run it yourself

Same product, two places to run it. Choose by how sensitive the recording is.

Hosted, on this site

Nothing to install. Use the hosted workspace for recordings that are not sensitive: your file is uploaded to our server and transcribed there, under the hosted privacy policy. Files from the free no-account preview are deleted automatically after 24 hours.

Self-hosted, on your hardware

Recordings, transcripts and search stay on infrastructure you control. Best for recordings under legal privilege, medical confidentiality or internal security policy.

How self-hosting works

Start free

Try one file with no account: up to 15 minutes and 25 MB. A free account keeps your transcripts; paid plans add minutes.

  • Free

    $0

    What every account gets without paying.

  • Pro

    $9.99/month

    For individuals.

  • Pro+

    $19.99/month

    For heavier and team use.

Compare plans

Free online tools

Simple tools for audio, video, images and documents. The image and PDF tools run in your browser where they can; see all tools.

Everything it does, in detail

Questions

Is it really free?
Yes, for one file at a time up to 15 minutes and 25 MB, with no account and no card. A free account lifts the per-file limits, keeps your transcripts in a library you can come back to.
Do I need to create an account?
Not to try it. Drop a file in and you get a transcript. You need a free account to save it, to transcribe longer or larger files, or to use speaker labels and search across everything you have transcribed.
What file formats can I upload?
Most video and audio containers, including MP4, MOV, MKV, WebM, AVI, MP3, WAV, M4A, FLAC and Opus. If a normal media player opens it, it will almost certainly work.
Can I paste a YouTube link?
Not yet. Download the video first, then upload the file — the YouTube transcript page walks through it, including how to use YouTube's own transcript panel when the video is not yours.
What languages does it handle?
33 languages, plus automatic detection when you are not sure. It can also translate speech in another language into English text.
How accurate is it?
Accurate enough that editing is faster than typing, and not accurate enough to publish unread. Clear speech and good audio come out close to clean; heavy accents, crosstalk, background noise and specialist vocabulary are where mistakes cluster. Every transcript is editable line by line for exactly that reason. We do not publish an accuracy percentage because the honest answer depends entirely on your audio.
What happens to my file?
It is uploaded to the Scriptorium server, transcribed there, and kept with your account until you delete it. Files from the free no-account preview are deleted automatically after 24 hours. See Privacy for the detail.
What is TryScriptorium?
A web app that turns speech in audio and video files into editable, timestamped text, with speaker labels, search and export. See the fact sheet for languages, limits and prices.
How is it different from YouTube's built-in transcript?
YouTube offers a transcript only for videos that have captions and you cannot edit it there. TryScriptorium transcribes any audio or video file you have, lets you correct every line, and exports SRT, VTT or text.
Does it support Arabic?
Yes. Arabic is a supported language and the site has an Arabic interface; see Arabic transcription.
Does it have an API?
No. A public API is not available today.
Where can I see measured speed and accuracy?
On the speed and accuracy page, with the method and hardware stated.

Try it with one file

No account, no card. See what the transcript looks like before you decide anything.

Transcribe free — no sign-up Create free account