Skip to content
Guides

How to transcribe an audio file to text for free

The short answer

Upload the file to a transcription tool, wait a few minutes, then check and edit the text. Microsoft 365 subscribers can use Word's Transcribe (300 minutes of uploads a month). Developers can run OpenAI's open-source Whisper on their own computer. Acta transcribes uploaded recordings with speaker names and writes a summary and action items, free for 15 days.

Step by step

  1. 1

    Start with clear audio

    One speaker per microphone and little background noise make the biggest difference to accuracy. Common formats like MP3, M4A, WAV and MP4 work in most tools.

  2. 2

    Pick a tool

    Word's Transcribe if you have Microsoft 365, Whisper if you're comfortable with the command line, or Acta if you also want speaker names, a summary and action items.

  3. 3

    Upload the file

    Most tools process a one-hour recording in minutes. Keep the page open if the tool asks you to.

  4. 4

    Name the speakers

    Rename Speaker 1, Speaker 2 and so on to real names, so the transcript is easy to follow.

  5. 5

    Check and export

    Skim the transcript against the audio for names, numbers and jargon, then export it as a document.

Three free or low-cost ways compared

  • Microsoft Word Transcribe: upload .wav, .mp4, .m4a or .mp3 files; separates speakers; up to 300 minutes of uploaded audio a month with a Microsoft 365 subscription.
  • OpenAI Whisper (open source): install with pip install -U openai-whisper and ffmpeg, then run whisper audio.mp3 --model turbo. Runs on your own computer, so nothing is uploaded.
  • Acta: upload recordings and files on every plan, including the 15-day free plan; transcript with speaker names, summary and action items, in many languages.

Transcribe a file with Whisper

Needs Python and ffmpeg. The turbo model works well for English; use a multilingual model such as medium with --language for other languages.

pip install -U openai-whisper
whisper interview.mp3 --model turbo

# Another language:
whisper meeting.m4a --model medium --language Spanish

Common mistakes

  • Recording on a laptop microphone at the far end of a meeting room.
  • Sharing a transcript without checking names, numbers and technical terms.
  • Uploading confidential recordings to a tool your organisation hasn't approved.
With Acta

Acta transcribes uploaded recordings and live meetings with speaker names, then writes the summary and action items. Record future meetings directly with the desktop app or Chrome extension, without a bot.

Try Acta free for 15 days

Questions

Can Google Docs transcribe an audio file?

Not directly. Google Docs voice typing listens to your microphone as you speak; it doesn't take an uploaded audio file. Use a transcription tool for files.

What is the most accurate free transcription?

Accuracy depends more on the audio than the tool. Whisper's larger models are among the most accurate free options; clear audio with one speaker per microphone helps every tool.

Can I transcribe audio in other languages?

Yes. Whisper's multilingual models support many languages, and Acta transcribes meetings in many languages too.

How long does it take to transcribe an hour of audio?

With an online tool, usually a few minutes. With Whisper on your own computer, it depends on the model size and your hardware.

Record your next meeting without a bot

Acta writes the transcript with speaker names, the summary and the action items for you. No notetaker joins the call. Free for 15 days.

Acta desktop app for Windows and Mac

Records Zoom, Microsoft Teams or any calling app from your own computer. It opens by itself when a Zoom or Teams meeting starts, and your screen is never recorded.

Download the desktop app

Acta Chrome extension

One click from the meeting tab in Google Meet, or Zoom and Teams in Chrome. Nothing else to install, so it works on a work laptop where you can't add apps.