Microsoft Word has a built-in Transcribe feature that turns uploaded audio files or live recordings into editable text, complete with speaker labels when multiple people are talking. It’s part of Word’s Office Intelligent Services, and here’s exactly how to use it.
What You Need Before Starting
Transcribe isn’t available to everyone by default. Microsoft Word offers two built-in speech-to-text tools — Dictation and Transcribe — and both are part of Microsoft’s Office Intelligent Services, available to Microsoft 365 subscribers. Specifically, you’ll need a Premium Microsoft 365 account, since the transcription features aren’t available in the free or desktop-only versions. You’ll also want to sign in using Microsoft Edge or Google Chrome, as other browsers can cause issues with the feature.
Supported audio formats are MP3, WAV, M4A, and MP4. Clear, well-recorded audio also makes a real difference in how accurate the output ends up being.
Step-by-Step: Uploading an Existing Audio File
- Sign in to your Microsoft 365 account and open Word Online (or the desktop app) in a supported browser.
- Go to the Home tab, click the drop-down arrow next to Dictate, and select Transcribe.
- In the panel that opens on the right, choose your audio’s language from the drop-down menu.
- Click Upload audio.
- Find and select your audio file in the file browser, then click Open.
- Wait while Word processes the file — you’re free to switch tabs or applications while it works.
Once finished, the transcript appears in the same pane, with each session including a timestamp, speaker name, and transcribed text. When Microsoft detects multiple speakers, it automatically separates the text by who’s talking.
Recording Audio Directly in Word (No File Needed)
If you don’t already have an audio file, Word can record and transcribe on the spot:
- Follow the same steps to open the Transcribe pane.
- Select Start recording instead of Upload audio.
- Allow Word to access your microphone and start talking. You can pause and resume recording anytime by clicking the microphone icon.
- When you’re done, click Save and transcribe now.
- Wait for Word to process and transcribe the recording.
Editing Your Transcript
Automated transcription isn’t perfect, so reviewing the output matters. Word is generally accurate at capturing all the words in an audio file, but some words may come through incorrectly, so it’s worth reading over the text before saving.
To fix an error, hover over the section of text that needs correcting — a pencil icon appears, and clicking it lets you edit that section directly in the pane.
A Few Quirks Worth Knowing
- Audio files are automatically saved to OneDrive, specifically in a /Transcribed Files folder, once the transcription completes.
- You can’t transcribe a file directly from OneDrive — it has to be saved or synced locally first, then re-uploaded to Transcribe.
- Starting a new transcription deletes your current one. Clicking New transcription in the lower right of the Transcribe window will delete the existing transcript sections, so make sure you’ve saved or copied anything you need first.
- This same workflow applies to video files too — if you’re working with video, you can extract the audio component and run it through the same transcription process.
Common Mistakes to Avoid
- Trying to use Transcribe on a free or desktop-only Microsoft 365 plan — the feature requires a Premium subscription.
- Uploading low-quality or noisy audio and expecting clean output — accuracy drops noticeably with poor recordings.
- Starting a new transcription without saving the current one first, since Word warns you but the deletion is otherwise irreversible.
Join The Discussion
Have you used Word’s Transcribe feature for meetings, interviews, or lectures, and how accurate did you find it compared to other transcription tools you’ve tried? Share any tips you’ve picked up for improving accuracy — audio quality tweaks, language settings, or editing shortcuts — and if you’ve run into the OneDrive upload quirk or lost a transcript by starting a new one, that’s a useful heads-up for others to hear about.
