Audio File Translator

How it works

From a recording in one language to text in yours, in three steps.

1. 1

   Upload the recording

   Drop in an MP3, M4A, WAV, MP4 or WebM file up to 10 minutes. Leave the language on Auto-detect or pick the one spoken, sign in with Google, Apple or an email code, and press Start Transcription.

2. 2

   Read the transcript and pick a language

   The transcript appears in the language it was spoken. Fix a misheard name, trim what you do not need, then choose the language to translate into; English is preselected on this page.

3. 3

   Translate, read it or hear it

   Press Translate. The translation appears with a note on any idiom and the original transcript kept under it; copy it, or press Read aloud to hear it in your browser's voice for that language.

What you get

The transcript as spoken, the translation as meant.

• The transcript in the original language

  Punctuated text of the recording as spoken, saved to your account, so you can check names, numbers and what exactly was said.

• A faithful translation

  The whole transcript, nothing added and nothing dropped, with names, numbers, dates, prices and units kept as they were said and the paragraphs kept.

• Notes on idioms

  When a saying does not translate word for word, the translation carries the meaning and a one-line note gives the original phrase, what it means literally and how it was rendered.

• 20 target languages

  English, Spanish, German, French, Portuguese, Italian, Dutch, Hindi, Japanese, Korean, Mandarin Chinese, Russian, Turkish, Polish, Swedish, Danish, Norwegian, Finnish, Indonesian and Ukrainian.

• Read aloud

  When your browser has a voice for the target language, a button speaks the translation, so you can play it to the person in front of you.

• Copy

  One button copies the translation and the notes as plain text for a message, a document or your CRM.

Translate audio to text online: how to get a good result

Two things decide the result: how well the recording was transcribed, and how the transcript was translated.

• Pick the spoken language when you know it

  Auto-detect works, but choosing Spanish, German or whichever language was spoken usually gives a cleaner transcript, and the translation can only be as good as the transcript.

• Check names and numbers before you translate

  People's names, product names, prices and dates are the words most often misheard. Fix them in the transcript box; the translation follows that text.

• Trim to the part you need

  A 10-minute transcript is around 1,500 words. Translating only the part that matters keeps the notes useful and the result quick to read.

• The register is kept

  Formal speech stays formal and casual speech stays casual in the target language; questions stay questions.

• It is a translation, not a legal or medical record

  Use it to understand what was said. For a contract, a consent form or anything a doctor or a lawyer will rely on, use a professional translator.

Recordings people translate here

A recording in one language, understood in another.

• A Spanish voicemail or voice note

  A customer, a crew lead or a supplier left a message in Spanish: upload it, read the Spanish, translate it to English, and answer knowing what was asked. Spanish to English audio translation is the most common job on this page.

• A bilingual meeting

  A call with the office in Mexico, Madrid or Berlin: transcribe it in the language most of it was in and translate the transcript into yours.

• A foreign-language interview

  A journalist or researcher with an interview in another language gets the transcript to quote from and a translation to work with; the notes show where a literal rendering would mislead.

• A supplier's or a distributor's message

  Order details, delivery dates and prices in a voice message from abroad, translated with the numbers kept exactly as said.

• A training video in another language

  The converter takes MP4 and WebM too, so a recorded webinar or demo in another language becomes a transcript and a translation: translate video to text, up to 10 minutes.

A passage, a recording, or every call

Three tools for three sizes of job; all start free.

• Spanish to English Voice Translator

  Live: speak Spanish into the microphone or paste up to 2,000 characters and read the English on the spot, or the other way round. No recording needed.

• This page: an uploaded recording

  A file up to 10 minutes in any of the converter's 23 languages: the transcript in the original language, then a translation into one of 20 languages, with the original kept next to it.

• The ParrotNotes app: every call and meeting

  On the Pro plan ParrotNotes transcribes every recording in 99+ languages and gives you the transcript in the language you choose, so a bilingual meeting lands in your language with a summary and action items, with no upload step.