Transcribe audio and video to text
This free audio and video transcription tool turns speech into text, word by word, with speaker names and timestamps. It’s made for interviews, meetings, lectures, podcasts, webinars, voice memos, court and research recordings, and the sound of any video.
There are four ways in: upload files (pick several and they’re transcribed one after another), transcribe a playing tab (a video, meeting or stream playing in another tab), paste a link to a video page, a file or a podcast feed (video pages play right in the tool and are transcribed while they play), or record with your microphone. The words appear while the rest is still being transcribed.
Everything runs in your browser. Your recordings are never uploaded, and there’s no account, no watermark and no limit on minutes or on how many files you transcribe.
In the tool, How to use in the top bar opens a short guide, and the search box finds any option: type what you want to do, like “rename a speaker” or “subtitles”, and it takes you there.
How to transcribe audio to text
- Choose the settings. Pick the language (or leave Detect automatically), the accuracy, and how many people are speaking.
- Add your recording. Drop audio or video files on the page (MP3, WAV, M4A, OGG, FLAC, MP4, MOV or WebM, up to 6 hours each), or use one of the other three ways in.
- Read along while it works. The text fills in part by part. When it’s done, the speakers are found and each paragraph gets a name.
- Fix and name. Type real names for Speaker 1, Speaker 2 and so on, double-click a paragraph to correct a word, and click any word to play the recording from there.
- Download. Click Download and save the transcript as Word, PDF, TXT, Markdown, SRT, VTT, CSV or JSON, or copy the text.
Four ways to add a recording
- Upload files: drop one file and it opens right away. Drop several and they wait in line under Your transcripts, each with its own progress bar, so you can keep working.
- Transcribe a playing tab: play a video, an online meeting, a webinar or a live stream in another tab, then choose that tab and turn on Also share tab audio. The words appear about 20 seconds behind, and the sound is recorded so you can play it back with the transcript. Turn on Also record my microphone to include your own voice in a meeting. This works in Chrome and Edge on a computer.
- Paste a link: paste the link of a video page and it plays right in the tool. Click Play here and transcribe, click Allow when your browser asks to share the tab and keep Also share tab audio on, and the words appear while it plays. When the video ends, the transcript finishes by itself. A page that can’t play in the tool opens in a new tab: start the video there, come back and click Choose that tab. You can also paste a direct link to an audio or video file, a web page with a player, or a podcast feed (RSS) to pick an episode. Playing a video page needs Chrome or Edge on a computer.
- Record with your microphone: pick your microphone, check the level bar, and talk. The text appears while you speak, and the recording is saved with it.
Speaker labels
Speakers are found automatically, so a conversation reads like a script: Speaker 1, Speaker 2, Speaker 3, each with its own color. If you know how many people are talking, choose 2 to 6 speakers for the best result, or One speaker for a lecture or voice memo.
- Rename a speaker: type a name in the Speakers card and it changes everywhere.
- Fix who said what: click the name above a paragraph and pick the right person.
- Merge two speakers: when one voice was split in two, merge it into the other.
- Talk time: see how long each person spoke, in minutes and as a share of the conversation.
- Find speakers again: change the number of speakers and only the speakers are found again. The text and your corrections stay.
Edit, search and play along
- Click a word to play the recording from that moment. The spoken word is highlighted as it plays, and the text follows along.
- Double-click a paragraph to correct it. Press Ctrl+Enter to save or Esc to cancel.
- Find and replace: press Ctrl+F to find a word, then replace one match or all of them, for example a name the speech model misheard.
- Playback speed from 0.5× to 2×, skip back and forward 5 seconds, and a timeline colored by speaker.
- Undo and redo every change with Ctrl+Z and Ctrl+Shift+Z.
Download Word, PDF, subtitles and more
- Word (.docx) and PDF: a clean document with the title, length, speaker names and times, ready to share or print.
- Text (.txt) and Markdown (.md): plain text, with or without speaker names and times.
- Subtitles (.srt and .vtt): cues of up to two lines, split at sentence ends, pauses and speaker changes. Choose short, normal or long lines, and whether to show speaker names.
- Spreadsheet (.csv): one row per paragraph with start, end, speaker and text.
- JSON: every word with its start and end time, for your own apps.
- Recording: download the sound of a tab or microphone recording, or the original file.
Languages and translation
Transcription works in 60 languages, including English, Spanish, French, German, Italian, Portuguese, Dutch, Hindi, Arabic, Russian, Ukrainian, Polish, Turkish, Japanese, Korean, Chinese, Vietnamese, Indonesian, Bengali, Tamil, Urdu and Persian. Leave Detect automatically on and the language is found from the first words.
Turn on Translate to English to get English text from a recording in any of those languages.
Accuracy
Choose how accurate the transcript should be. The speech model is downloaded once and kept by your browser:
- Fast: about 40 MB, quickest on phones and older computers.
- Balanced: about 80 MB, a good choice for most recordings.
- Most accurate: about 250 MB, best for noisy sound, accents and names.
You can change the accuracy or the language later and click Transcribe again.
Your recordings are private and saved
Transcription and speaker labels run on your own device, so your recordings never leave it. That makes it a good fit for confidential interviews, patient notes, legal recordings and research data. Your transcripts are saved in this browser as you work, under Your transcripts on the start screen, so you can close the tab and come back later in the same browser.
Tips for better transcripts
- Record close to the microphone in a quiet room. Less echo and background noise means fewer mistakes.
- Set the number of speakers when you know it. It helps most when two voices sound alike.
- Choose the language instead of Detect automatically when a recording starts with music or silence.
- Use Most accurate for names, technical words, strong accents or noisy sound.
- Fix names with Find and replace in one step instead of paragraph by paragraph.
Frequently asked questions
Is this transcription tool free?
Yes. It’s free, with no sign-up, no watermark and no limit on minutes or on how many recordings you transcribe.
Are my recordings uploaded?
No. Transcription and speaker labels run in your browser, and your recordings stay on your device. Only the tool’s own files are downloaded, such as the speech model and the speaker model.
Does it label the speakers?
Yes. Each paragraph is labeled Speaker 1, Speaker 2 and so on. Type real names in the Speakers card to rename them everywhere, and click the name above a paragraph to fix who said it.
Can I transcribe a video?
Yes. Upload an MP4, MOV or WebM file and its sound is transcribed. The video plays in the side panel, and clicking a word jumps to that moment.
Can it transcribe a video from a website or an online meeting?
Yes. Paste the video’s link, click Play here and transcribe and click Allow with Also share tab audio on. The video plays in the tool and the text appears while it plays. For an online meeting, or a page that can’t play in the tool, play it in another tab, click Transcribe a playing tab and pick that tab. This works in Chrome and Edge on a computer.
Can I paste a link instead of uploading?
Yes. A video page plays right in the tool and is transcribed while it plays. A direct link to an audio or video file, a web page with a player or a podcast feed is downloaded and transcribed right away when the site lets other websites read its files. When it doesn’t, the file plays in the tool and is transcribed while it plays, or you can save it and upload it, which is faster.
Which languages are supported?
60 languages, including English, Spanish, French, German, Hindi, Arabic, Portuguese, Japanese, Korean and Chinese. The language can be found automatically, and any of them can be translated to English.
Can I make subtitles?
Yes. Download SRT or VTT subtitles timed to the words, with up to two lines per cue. Choose the line length and whether to show speaker names.
How long can a recording be?
Up to 6 hours per file. Longer recordings take more time, but the words appear part by part, so you can start reading and editing early.
How accurate is it?
Clear speech close to the microphone comes out very accurate. Choose Most accurate for noisy recordings, strong accents or many names, and fix any misheard word with Find and replace.
Which browsers work?
A recent version of Chrome, Edge, Firefox or Safari on a computer works best. Transcribing a playing tab needs Chrome or Edge on a computer. It also works on phones and tablets, though long recordings take longer there.
