Convert any audio recording into accurate text online — no install, no typing.
99 languages, timestamps, export to TXT, DOCX, PDF or SRT. Free to start.
The link must be publicly accessible.
A free online converter that does the listening for you — drop in a file and get editable text back, all in the browser.
Upload an MP3, WAV, M4A, AAC, OGG or FLAC recording — or a video file, or a link to one on YouTube, TikTok, Instagram, Facebook, X, Vimeo, Telegram, Google Drive or Dropbox. Nothing to download and no account needed to start.
The recording is queued as a job, the language is read from the audio, and the speech comes back as punctuated text broken into timed segments. There are no settings to fiddle with because there is nothing useful to set.
Edit against the player if anything needs fixing, then export to TXT, Markdown, DOCX, PDF, or SRT and VTT subtitles. The transcript is yours to keep, and it stays in your history even after the audio is gone.
This isn't a format swap — it listens to the recording and writes out what was said, ready to use.
Skip the hours of manual typing — a long recording is split into chunks that convert at the same time, so the wait is minutes rather than the length of the audio.
Interviews, calls, memos, lectures, field recordings — if there is speech in it, the converter writes it out, video files included.
Clean punctuated text you can copy, search, translate or paste anywhere, with each phrase still tied to the moment it was spoken.
Nothing to install on Windows, Mac, Linux, Android or iOS — the page hands your browser a signed URL and the file goes straight to storage.
Speech in almost any language converts without a language picker; the spoken language is detected from the recording itself.
Download as TXT, Markdown, DOCX, PDF, or SRT / VTT subtitles built from the same timed segments.
Desktop transcription apps are slow to set up and tie you to one machine. An online converter just works: open the page, drop in a recording, and get accurate text back in 99 languages with timestamps and punctuation. SlayScribe runs the transcription for you with OpenAI Whisper, so the output is editable and exportable the moment it's done — no installs, no licences, no manual typing.
The parts of an online converter that usually go unexplained — and the ones that bite.
The page asks for a signed upload URL and your browser sends the file straight to storage with it. That matters for a practical reason: a serverless request body tops out around 4.5 MB, which is about three minutes of decent-quality audio, and every converter that posts the file through its own API inherits that ceiling. Guests get the same path — the key they upload to is signed, so no account is needed to move a large file.
Every upload becomes a row in a queue before any transcription starts, and that row is the state of the job. Navigate away, reload, close the browser, run out of battery: the work carries on and the result is waiting in your history. This is not decoration. When short files used to be transcribed inside the request instead, the progress existed only in the tab, so a walk to the pricing page and back showed an empty list — and re-uploading the same file spent a second helping of your minutes for a transcript that was already being made.
Paste a link and there is nothing to upload: the media is fetched server-side from YouTube, TikTok, Instagram, Facebook, X, Vimeo, Telegram, Google Drive or Dropbox, then converted the same way as an upload. What a link cannot do is get past a login — a private post, an unlisted file with restricted sharing or an age-gated video will fail on the fetch, not on the transcription, and the honest fix is to download it yourself and upload the file.
The transcript — text, segments, timestamps — is kept for as long as your account exists. The audio behind it is not: on the free plan and for unclaimed guest transcripts, the stored recording is deleted 30 days after the transcript is made, which keeps idle free accounts from turning into a storage bill. A paid account keeps its audio indefinitely, because the player next to the text is a part of what it pays for.
Three different ceilings, and the message tells you which one you hit. Above 5 GB the upload is refused outright. A file over a quarter of a gigabyte is routed to the background worker instead of being handled in the page — signed in, that is invisible to you; as a guest it is declined, because there is no account to attach a long-running job to. And past the per-file length ceiling (60 minutes free, 10 hours on Pro) the file is refused after the duration is probed, before any minutes are spent.
Converting audio to text is free within a monthly allowance of 60 minutes, with no card and no account required to start; Pro raises the allowance, lifts the daily file limit and allows much longer recordings.
The converter runs entirely in the browser on any device — Windows, Mac, Linux, Android or iOS — because the transcription happens on our side and the page only needs to upload the file and show the result.
MP3, WAV, M4A, AAC, OGG and FLAC are the common ones, along with video files and links; anything ffmpeg can decode is converted to a mono audio track before transcription, so unusual containers usually work too.
The converter uses OpenAI Whisper, and the result tracks the recording rather than the plan: clean speech converts close to verbatim, while background noise, overlapping voices and unfamiliar names are where corrections are needed. The text sits beside the player so fixing them is a couple of minutes.
The transcript downloads as TXT, Markdown, DOCX, PDF, SRT or VTT, all built from the same timed segments, so the subtitle files line up with the audio without any further work.
The conversion carries on, because SlayScribe queues every upload as a job row before transcription starts. That row survives navigation, a reload, a closed tab and a crashed browser, so the finished TXT, DOCX or SRT is waiting in your history when you come back.
Paste a link from YouTube, TikTok, Instagram, Facebook, X, Vimeo, Telegram, Google Drive or Dropbox and the media is fetched on our side, then converted like an upload. Links that require a login or an age check cannot be fetched — download those yourself and upload the file.
Up to 5 GB per upload, within 60 minutes of audio on Free and 10 hours on Pro. Files above roughly a quarter of a gigabyte are handled by the background worker, which needs an account — as a guest, sign in first.
Free and unclaimed guest recordings are deleted 30 days after the transcript is made, while the transcript itself — text, segments and timestamps — stays in your history. Paid accounts keep the audio, so the player beside the transcript keeps working.
Speech in 99 languages, detected automatically from the audio, with no language to pick before you start and nothing to change afterwards.
Go Pro to transcribe hours of audio and video, with more monthly minutes and unlimited uploads — accurate text in 99 languages.
See Pro plans