English Transcription

Transcribe English audio to text, free

Upload a English recording and get a clean, readable transcript in seconds. Punctuation and sentence casing are inferred, so you get paragraphs rather than an unbroken lowercase stream. Filler words are kept — delete them yourself if you want a clean read. Free to start, no account required.

New uploads are paused until it finishes, so the service stays fast for everyone. You can leave this page — the result will be waiting in your queue.

Open my queue →

MP3, M4A, WAV, FLAC, OGG, MP4 · up to 20 MB as a guest

Transcribing your audio…

How accurate is English transcription? Tested

What to expect

Quick read-through

Most sentences come out right; skim for names and numbers.

We ran 20 English recordings from Google's FLEURS dataset — native speakers reading everyday news-style sentences — through the same Whisper large-v3-turbo setup this site uses, with English selected, on 30 October 2026, and compared every transcript with what was actually said.

A typical sentence from the test · underlined words differ from what was said

Said: To the north and within easy reach is the romantic and fascinating town of Sintra and which was made famous to foreigners after a glowing account of its splendours recorded by Lord Byron.

Transcript: To the north and within easy reach is the romantic and fascinating town of Cintra and which was made famous to foreigners after a glowing account of its splendorous recorded by Lord Byron.

Clear read speech is an easy case. Interviews, phone calls, several people talking at once and music under speech produce more errors. FLEURS sentences © Google, CC BY 4.0. How we measured · All 40 languages compared

What makes English hard to transcribe

English has more training data behind it than any other language, so raw accuracy is rarely the problem. What trips transcription up is accent spread: an Indian, Scottish and Nigerian speaker in one meeting are further apart phonetically than many separate languages are from each other.

What your English transcript looks like

Punctuation and sentence casing are inferred, so you get paragraphs rather than an unbroken lowercase stream. Filler words are kept — delete them yourself if you want a clean read.

Timestamped output English
[00:00 → 00:04] Thanks everyone for joining. Let's start with last month's numbers.
[00:04 → 00:08] I'll share my screen so you can follow along.

An illustration of how English transcripts are formatted — script, punctuation and timestamps. Not the output of any particular recording.

How accurate to expect

English is one of the strongest languages — clear recordings usually need only a quick read-through, and Spanish, Italian and German measured just as well. The practical limits are overlapping speakers and heavy background noise rather than the language itself.

What usually needs fixing

  1. 1 Industry acronyms and product names — the model hears the sounds correctly but has no way to know your internal spelling.
  2. 2 Speaker overlap in group calls, where two voices in the same second collapse into one line.
  3. 3 Homophone pairs in legal and medical contexts, where 'affect' and 'effect' carry real consequences.

Getting better results

  • → Technical acronyms are the most common error source; a glossary pass beats re-recording.
  • → For interviews with two distinct accents, transcribe the whole file at once rather than splitting — context helps the model.

Who transcribes English

Podcasters producing show notes and searchable episode archives
Research teams coding qualitative interview data across multiple studies
Newsrooms turning press conferences into quotable copy under deadline

Questions about English transcription

Yes, though accuracy varies. Broad Scottish, Nigerian and South Asian English are all recognised, with proper nouns being the weakest spot.
Every transcript comes with both a clean version and a timestamped version, so you can cut subtitles from the same job.
Not into labelled speakers. You get a continuous transcript with timestamps, which most editors find faster to clean than mislabelled diarisation.
Up to 30 minutes per file on the free plan and 90 minutes on Pro. Longer recordings should be split at natural breaks.
Yes — choose English and upload as usual. Our accuracy test used American speakers; strong regional accents mostly add errors in names, place names and fast speech, so give those a closer read.

Need the opposite — English text to speech?

Turn a English script into a voiceover with 75 English AI voices — preview each one free in your browser.

English voices →

Transcribe other languages

Languages that also use the Latin alphabet are listed first.

SpeakScribe transcribes 32 languages with dedicated guidance, and 99 in total via auto-detect. See all audio-to-text options →

Transcribe your English audio now

Upload a file and get a timestamped transcript in seconds — free, no account needed.

Open the transcriber