How to Transcribe Audio to Text in Word

Yes, Microsoft Word can transcribe a recording or an audio file, provided you have access to the Transcribe feature in Microsoft 365. Open a document, go to Home > Dictate > Transcribe, then choose Upload audio for an existing file or Start recording for a live conversation.

The important detail is that Dictate and Transcribe are different tools:

Dictate turns your voice into text in the document while you speak.
Transcribe records or uploads a complete file, separates the result into timestamped speaker sections, and lets you review the transcript before adding it to the document.

This guide covers both Word workflows, the current Microsoft 365 requirements, what to do when Transcribe is missing, and how to turn the first draft into a Word document someone can actually read and use.

Before you start: Word Transcribe requirements

Check these details before looking for the Transcribe button:

Requirement	Current Word Transcribe support
Account	A signed-in Microsoft 365 account with access to Transcribe
Reliable route on Mac or an unsupported desktop app	Word for the web in a current Edge or Chrome browser
Document command	Home > Dictate dropdown > Transcribe
Upload formats	WAV, MP3, M4A, and MP4
Uploaded-audio allowance	300 minutes per month for Microsoft 365 subscribers; 30,000 minutes with a Microsoft Copilot license
File storage	Uploaded and recorded files are placed in the Transcribed Files folder in OneDrive

Microsoft currently documents the Windows desktop feature for Word for Microsoft 365 in commercial tenants. Government tenants are directed to Word for the web. If you use Word on a Mac, or the desktop command is absent, open the document in Word for the web instead of assuming your installation is broken.

These limits can change. Microsoft's current Word Transcribe documentation is the source to check before planning a large batch. It confirms the formats and monthly uploaded-audio allowance above, but does not currently state a single-file size limit in the article. Be wary of old tutorials that present an unreferenced MB limit as a current rule.

How to upload an audio file to Word

Use this method when the meeting, interview, lecture, voice memo, or podcast is already recorded.

Sign in to Microsoft 365 and open a blank or existing Word document.
Select Home, open the dropdown below Dictate, and choose Transcribe.
In the Transcribe pane, select the spoken language that matches the recording.
Select Upload audio.
Choose a WAV, MP3, M4A, or MP4 file from your computer.
Keep the Transcribe pane open while Word uploads and processes the file. You can work in another document or browser tab, but do not close the pane.

Processing is not always instant. Microsoft notes that it can take up to about the length of the recording, depending in part on your internet connection. When it finishes, the pane displays timestamped sections and labels detected voices as Speaker 1, Speaker 2, and so on.

An MP4 is accepted because Word transcribes the spoken audio inside the video; it does not convert the visual content into a written description.

How to record and transcribe directly in Word

Use direct recording when you want to capture a conversation and take notes in the same document.

Go to Home > Dictate > Transcribe.
Check your computer's microphone input before the conversation begins. A headset microphone selected after you put the headset aside, for example, can produce a poor or empty recording.
Select Start recording. On first use, your browser may ask for microphone permission.
Wait until the microphone button has a blue outline and the timer starts, then begin speaking.
Leave the Transcribe pane open. Use the microphone button to pause and resume when necessary.
Select Save and transcribe now when the conversation ends.

Word saves the recording to the Transcribed Files folder in OneDrive and then creates the transcript. Text does not appear live in the document during this process. If you need words to enter the document as you speak, use Word Dictate instead.

For an in-person meeting, place the microphone where it can pick up everyone without rubbing against a desk or clothing. For an online call, make a short test recording first. Capturing only your headset microphone can omit the other participants.

Review the transcript before adding it to Word

A completed transcription is a draft, not a finished record. Word gives you several useful review controls in the Transcribe pane:

Select a timestamp to play that section of the recording.
Follow the highlighted section during playback and adjust playback speed, up to 2x.
Hover over a section and select the edit control to correct its text.
Change Speaker 1 to a person's name and apply that name to every matching speaker label.
Use Add section to document for one passage, or Add all to document for the complete transcript.

Transcribe does not automatically put its result into the document. The transcript remains associated with that document when you close and reopen it, but Microsoft allows only one stored transcript per document. Starting a New transcription removes the current transcript from the pane. Sections you already added to the document remain as ordinary document content.

The audio link and playback access deserve separate attention when sharing. Adding the transcript to the document does not automatically give every recipient permission to play the OneDrive recording. Share the audio file separately only when recipients need it and are allowed to hear it.

Turn the raw transcript into a useful Word document

Automatic transcription saves typing time; editing is what makes the result useful. Work through the document with a purpose instead of correcting every line in the same way.

Verify high-risk details against the recording

Replay the source whenever the text contains:

names, job titles, company names, or locations;
dates, times, prices, quantities, percentages, or measurements;
technical terms, acronyms, product names, and medications;
direct quotations, decisions, deadlines, or commitments;
speech affected by background noise, overlap, or a distant microphone.

Do not treat an unreviewed transcript as an official meeting record, certified transcription, or the sole source for an important legal, medical, employment, or financial decision.

Format for the reader, not for the transcription engine

Use actual names when readers need to follow a multi-person conversation. For a single-speaker lecture or voice memo, repeating a speaker label before every paragraph usually adds clutter.

Combine short fragments into logical paragraphs. Add headings when the topic, agenda item, interview question, slide, or chapter changes. A one-paragraph summary and an action-item list can make a long meeting transcript easier to navigate, provided you clearly separate those additions from what was actually said.

Keep timestamps when they help a reader return to the source—for example, at a topic change or an important quotation. Delete granular timestamps when they interrupt normal reading. The W3C Web Accessibility Initiative transcript guidance similarly recommends logical paragraphs and sections, useful headings, and timestamps only where they add value.

For accessibility, a basic audio transcript should include meaningful speech and non-speech audio needed to understand the content. A transcript derived from video may also need important visual information that nobody says aloud, such as a speaker's displayed name, essential on-screen text, or a demonstrated action. Word can transcribe the audio track, but it cannot decide which visual details your reader needs.

Why is Transcribe not showing in Word?

If Home > Dictate does not contain Transcribe, check these causes in order.

1. You are signed out or using the wrong account

Confirm that Word is signed in to an active Microsoft 365 account. If you have both personal and work accounts, the open document may be using a different account from the one with the subscription or organizational permission.

2. Your Word version does not include this Transcribe path

Having a Dictate button does not prove that you also have Transcribe. Microsoft's separate Dictate documentation covers more desktop and perpetual versions, while its Transcribe documentation is narrower. Word 2016, 2019, 2021, or another one-time-purchase edition should not be assumed to include the Microsoft 365 Transcribe workflow.

3. You are using Word for Mac

Microsoft's current desktop Transcribe instructions identify Windows commercial tenants, not Word for Mac. Open Word for the web in Edge or Chrome, sign in, and try Home > Dictate > Transcribe there.

4. Your organization restricts connected experiences

Word desktop uses Microsoft's optional connected experiences for transcription. A company or school administrator may disable that service or make it unavailable for your tenant. If the web route and correct account do not help, ask your Microsoft 365 administrator whether Transcribe is allowed.

5. The browser or microphone is blocked

For Word for the web, use an up-to-date Edge or Chrome browser. Allow microphone access only if you are recording directly. Microphone permission is not the usual cause of a failed upload, so do not change it unnecessarily when you are only selecting a file.

6. The file format or monthly allowance is the problem

Word currently accepts WAV, MP3, M4A, and MP4 uploads. A file named with one of those extensions can still fail if its actual media encoding is damaged or incompatible, so try playing it locally or exporting a fresh MP3/WAV copy. Also check whether the account has used its monthly uploaded-audio allowance.

When Word is not the right transcription tool

Word is convenient when your file already matches its supported formats and the final destination is a Word document. A dedicated transcription workflow is more useful when you have another media format, need tighter speaker controls, want to guide recognition of specialist vocabulary, or also need subtitle files.

With Transcribe to Text's audio and video transcription workflow, you can:

Upload one supported audio or video file, up to 5 GB.
Choose the spoken language or use automatic detection.
Enable speaker identification and, when useful, provide an expected count or range.
Add brief context and key terms for names, brands, acronyms, or specialist phrases.
Review the transcript while replaying the source, correct individual segments, and assign clear speaker names.
Save the revision and export TXT or Markdown for a document. Open the TXT file in Word, or copy the reviewed Markdown text into Word and apply your document styles.

The current upload list includes MP3, WAV, M4A, AAC, FLAC, OGG, OPUS, MP4, MOV, AVI, MKV, MPEG/MPG, and WebM. SRT and VTT exports are also available when the same transcript will be used for captions or subtitles. Transcription requires an account and sufficient credits; it is not presented as an unlimited free service.

Choose Word Transcribe when…	Choose a dedicated workflow when…
The recording is WAV, MP3, M4A, or MP4	The media uses another supported audio or video format
You already have an eligible Microsoft 365 environment	Word Transcribe is missing or its upload allowance is exhausted
The transcript only needs to become Word content	You need TXT, Markdown, SRT, or VTT outputs
Basic automatic speaker separation is enough	You want speaker-count controls, context, and key terms before processing
Storing the recording in OneDrive fits your workflow	You want a review workspace centered on synchronized media and saved corrections

Whichever tool creates the draft, the last step is the same: replay uncertain passages and edit for the person who will rely on the document.

Frequently asked questions

Can Microsoft Word transcribe audio for free?

Word Transcribe is tied to eligible Microsoft 365 access rather than offered as an unlimited standalone free transcription tool. Microsoft currently lists 300 minutes of uploaded audio per month for Microsoft 365 subscribers and 30,000 minutes for users with a Microsoft Copilot license.