Skip to content
Transcription

How to transcribe audio in Microsoft Word: the steps, the limits and what to do when you hit them

Updated 10 min read

Short answer

To transcribe audio in Word, go to Home > Dictate > Transcribe, then record or upload a .wav, .mp4, .m4a or .mp3 file. Microsoft says the feature needs a Microsoft 365 subscription and caps uploaded audio at 300 minutes a month. For long files, YouTube links or SRT captions, upload the recording to FilmScribe instead.

Step by step

  1. 01 Sign in with your Microsoft 365 subscription Open Word for the web in the new Microsoft Edge or Chrome and sign in with the account that holds the active subscription. Without it, the button may not show up.
  2. 02 Open the Transcribe pane On the Home tab, open the Dictate dropdown and choose Transcribe.
  3. 03 Record or upload the audio Select Start recording to record on the spot, or Upload audio to send a .wav, .mp4, .m4a or .mp3 file.
  4. 04 Wait with the pane open Transcription can take up to about the length of the audio. Keep the Transcribe pane open while it runs.
  5. 05 Review and add it to the document Fix the text, replace Speaker 1 with the real name and select Add all to document, or add only the sections you need.
  6. 06 Past what Word takes? Upload to FilmScribe A long file, video in another format, a YouTube link or SRT captions: transcribe in FilmScribe and open the exported TXT in Word.
In this guide
  1. The short version
  2. Can Microsoft Word transcribe audio?
  3. How to transcribe audio in Word, step by step
  4. Transcribe vs Dictate in Word: what’s the difference?
  5. How to edit the transcript and add it to your document
  6. Where Word Transcribe stops
  7. How to transcribe in FilmScribe what doesn’t fit in Word
  8. Word or FilmScribe: which one for which job
  9. How to get a FilmScribe transcript into Word
  10. What if the recording is a video or a YouTube link?
  11. What affects accuracy, in Word or anywhere else
  12. Other options: OneNote, YouTube and your phone’s recorder
  13. Common mistakes
  14. Frequently asked questions
  15. Sources

Yes, Microsoft Word can transcribe audio. The feature is called Transcribe, it sits inside the Dictate button, and it turns a recording into text with speakers separated and a timestamp on each section. For a short interview or a meeting whose notes will live in the same document, it does the job.

What decides whether it works for you is the fine print. Transcribe needs a Microsoft 365 subscription, has a monthly cap on minutes and takes only four file formats. This guide walks through the steps as Microsoft’s documentation describes them (we read it on November 11, 2026), says plainly where the feature stops, and shows how to transcribe everything past that point in FilmScribe and still finish the text in Word.

The short version

  • In Word, go to Home > Dictate > Transcribe. You record on the spot or upload a .wav, .mp4, .m4a or .mp3 file.
  • Microsoft says the feature needs a Microsoft 365 subscription and takes up to 300 minutes of uploaded audio a month. In Word for the web it only works in the new Microsoft Edge and Chrome.
  • Word separates speakers, timestamps each section and puts the text in your document. The documentation mentions no YouTube links, caption files or chapters.
  • For a long recording, video in another format, a YouTube link or SRT captions, upload to FilmScribe. We transcribe with a timestamp on every word and speakers separated.
  • Word is still a great place to finish: export TXT from FilmScribe and open it in Word.

Can Microsoft Word transcribe audio?

It can, under three conditions that Microsoft spells out on the feature’s support page.

Subscription. Transcribe is part of Word for Microsoft 365, and the page states its usage limits for subscribers. In its troubleshooting section, it tells anyone who can’t see the button to make sure they’re signed in with an active Microsoft 365 subscription.

Where it works. The page covers Word for the web and Word for Windows. In Word for the web, Transcribe only works in the new Microsoft Edge and Chrome, with an internet connection. In Word for Windows, Microsoft says the feature is currently available only for commercial accounts (“Commercial Tenants,” in its words), and that government tenants get it only in Word for the web. The page doesn’t mention Word for Mac or the phone apps. Microsoft itself suggests trying the feature in Word for the web.

Usage cap. With a Microsoft 365 subscription you can transcribe a maximum of 300 minutes of uploaded audio per month. With a Microsoft Copilot license, the maximum is 30,000 minutes per month.

What you want to knowWhat Microsoft states
Where it isHome tab, Dictate dropdown, Transcribe
What you needAn active Microsoft 365 subscription and an internet connection
Browser (Word for web)The new Microsoft Edge or Chrome
Monthly cap300 minutes of uploaded audio; 30,000 with a Microsoft Copilot license
Supported formats.wav, .mp4, .m4a and .mp3
SpeakersSeparated and labeled Speaker 1, Speaker 2; you can rename them
Languages80+ locales, including many English and Spanish variants
Where the audio is keptIn the Transcribed Files folder of your OneDrive
How many transcriptsOne per document

If you don’t have the subscription, or your recording is past those numbers, the answer is to transcribe audio to text outside Word and bring the result into your document. We show how further down.

How to transcribe audio in Word, step by step

The common case is a recording you already have: an interview on your phone, the audio of a meeting, a recorded class.

  1. Open Word and sign in. In Word for the web, use the new Microsoft Edge or Chrome and sign in to Microsoft 365.
  2. Open the Transcribe pane. On the Home tab, click the arrow next to Dictate and choose Transcribe. The pane opens beside your document.
  3. Upload the file. Select Upload audio and pick the recording. Supported formats are .wav, .mp4, .m4a and .mp3.
  4. Wait with the pane open. Microsoft warns that transcription can take up to about the length of the audio file, depending on your connection. You can work in another tab or app, but the Transcribe pane has to stay open.
  5. Check the result. The text shows up in the pane, split into sections, each with a speaker label and a timestamp. The audio file is saved to the Transcribed Files folder in your OneDrive.

To record straight into Word

If the conversation hasn’t happened yet, you can record inside the document and take notes on the page at the same time.

  1. Go to Home > Dictate > Transcribe and select Start recording. The first time, your browser asks for permission to use the microphone.
  2. Wait for the microphone button to be outlined in blue and the timestamp to start counting. That’s the sign that recording has begun.
  3. Speak clearly and leave the pane open. The same microphone button pauses and resumes the recording.
  4. When you’re done, select Save and transcribe now. Word saves the recording to OneDrive and starts transcribing.

Unlike dictation, no text appears on the page while you talk. You only see the transcript after you save.

Two tips from Microsoft make a real difference. Check which microphone your device is using: if the input is still set to the headset from your last call, an in-person conversation will record badly. And to record a virtual call, don’t wear your headset, so the recording can pick up the sound coming out of your device.

Transcribe vs Dictate in Word: what’s the difference?

They share a button and do different jobs.

  • Dictate types as you talk. You turn on the mic, speak, and the words land on the page, with voice commands for punctuation. It’s for writing. In the Word app, Microsoft says dictation is also only for Microsoft 365 subscribers.
  • Transcribe works from a recording, made on the spot or uploaded. The text comes afterward, split by speaker and timestamped, in a pane beside the document.

To turn an existing recording into text, Transcribe is the right tool. Playing the audio out loud next to your mic with Dictate on tends to go wrong: the room gets into the text, voices blend and you end up with no timestamps at all.

If what you wanted was to write by talking, see our guide on how to use voice typing and our speech to text tool.

How to edit the transcript and add it to your document

The transcript stays attached to the document until you remove it. Closing and reopening the pane, or the file, doesn’t delete anything.

  • Listen again. The controls at the top of the pane play the audio, and the section being played is highlighted. Click a section’s timestamp to hear just that part. Playback speed goes up to 2x.
  • Fix the text. Hover over a section, select Edit transcript section, make the correction and confirm.
  • Name the speakers. Word labels people Speaker 1, Speaker 2 and so on. While editing a section, replace the label with a name and use the option to change every instance of that speaker.
  • Move the text onto the page. Transcribe doesn’t write into the document on its own. Use Add section to document for one passage or Add all to document for the whole transcript. It arrives as regular text, with a link to the audio file.
  • Start another one. Each document stores one transcript. When you select New transcription, the current one is deleted from the pane; whatever you already added to the page stays.

To share, send the document as usual. For privacy, the people you share with can’t play the audio from the pane by default. If someone needs to listen, share the audio file as well, from the Transcribed Files folder in OneDrive.

As for your data, Microsoft says the audio is sent to it only to provide the service, and that once transcription is done the service doesn’t store the audio or the results. The recording stays in your OneDrive, where you can delete it.

Where Word Transcribe stops

The feature is good at what it sets out to do: a short recording, text in a document. The limits show when the job grows. Every point below comes from what Microsoft’s documentation says, or from what it doesn’t describe.

  • 300 minutes of uploaded audio a month. That’s 5 hours. Two two-hour recordings a month fit; a third won’t fit whole. The 30,000-minute maximum applies only with a Microsoft Copilot license.
  • Four formats. Only .wav, .mp4, .m4a and .mp3. WhatsApp voice messages (OGG or OPUS) and video in MOV, MKV or AVI are out and need converting first.
  • A recording or a file, nothing else. The two inputs described are recording in Word and uploading a file. The page describes no way to add a link, such as a YouTube video’s.
  • The pane stays open until it’s done. Transcription can take up to the length of the audio, and the pane has to stay open the whole time.
  • One transcript per document. A second recording means another document, or deleting the first one from the pane.
  • Section-level timing, text on the page. The pane timestamps each section, and the output described is text inside the document. The documentation mentions no caption file, such as SRT, and no chapters.
  • It depends on your account and browser. Without an active subscription, outside Edge and Chrome, or in Word for Windows without a commercial account, the button may not show up.

None of this is a flaw. Transcribe was built for people writing a document, not for people producing content from long recordings.

How to transcribe in FilmScribe what doesn’t fit in Word

In FilmScribe, the transcript is where the product starts. The workflow below covers the cases where Word stops: long files, video, YouTube links, SRT captions and chapters.

Audio as MP3, WAV, M4A, AAC, OGG and OPUS, video as MP4, MOV, AVI, MKV and WEBM, or a YouTube link: videos, Shorts, clips, youtu.be and livestreams, which we process once the replay is available. A file can be up to 10 GB and there’s no length limit; what counts is your plan’s hours. See the supported formats.

2. Mark only the part that matters

A three-hour meeting with one useful hour, a livestream with a long intro: set the start and end before processing. Only that range is transcribed, and only it counts toward your hours.

3. Get the text with speakers and a timestamp on every word

We transcribe word by word, in 90+ languages, with the language detected automatically and speakers separated on every plan. Once the upload finishes you can close the page: we email you when the file is ready. On Studio and Agency, you can name each speaker and have us work out who is who. See transcription and speakers.

4. Fix it in the text

Click any sentence and the player jumps to it. Correct names, acronyms and technical terms in place: the fix applies to the text on screen and to every export. Search finds any word across the whole file.

5. Export and finish in Word

The transcript exports as TXT, PDF and SRT, with the option to include speakers and times. Chapters by topic come ready in every file and can be copied in the format YouTube’s description reads. For your document, export TXT and open it in Word.

To try it on your own recording, the trial runs 3 days, includes 1 hour of material and asks for a card. There is no free plan.

Word or FilmScribe: which one for which job

SituationWord (Transcribe)FilmScribe
A short interview that becomes text in the same docWorks, if you already pay for Microsoft 365Works; export TXT and open it in Word
Recording and taking notes at the same timeWorks: it records inside the documentUpload the recording afterward
More than 5 hours of uploaded audio a monthPast the subscription’s 300 minutesNo length limit; your plan’s hours count
WhatsApp audio, video in MOV or MKVOutside the four formats; convert firstAccepted as they are
A YouTube video or livestream linkOnly a recording or an uploaded filePaste the link
SRT captions for YouTube or a video editorThe documentation mentions no caption fileSRT with TV, Netflix and YouTube presets
Chapters by topicNot mentioned in the documentationGenerated for every file
Finding the exact word in the audioA timestamp per sectionA timestamp per word; click a sentence to hear it
A name for each speakerYou rename Speaker 1, Speaker 2On Studio and Agency, you name them or we suggest names

The rule of thumb: if the recording is short, the text will stay in the document and you already pay for Microsoft 365, start in Word. If the recording is long, is video, lives on YouTube or will become captions, transcribe elsewhere and use Word only to finish.

How to get a FilmScribe transcript into Word

There are two ways, depending on what the document needs to show.

Text to edit. Export the transcript as TXT, with or without speakers and times, and open the file in Word (File > Open), or paste it into a new document. Apply heading styles to each topic, bold the names and save as .docx.

A table with the timing. If your reader needs the time of each line next to the text, for a review or a sign-off, export SRT and run the file through our free SRT to Word tool. It builds a .docx table with number, in time, out time and text, which opens in Word and Google Docs.

A transcript exported with speakers and times lands in your document like this:

[12:40] Person 1: What's the deadline for closing the report?
[12:44] Person 2: Friday. Only the cost spreadsheet is missing.
[12:52] Person 1: Then the review moves to Monday.

On Studio and Agency, the name you gave each person replaces Person 1 and Person 2. In Word, renaming Speaker 1 comes with Transcribe.

For academic work, our guide on how to transcribe an interview shows how to present quotes in the text. For the full walkthrough with any recording, see how to transcribe audio to text.

On Microsoft’s list, .mp4 is the only video format, and it goes in as an uploaded file. For everything else:

  • Video in MOV, MKV, AVI or WEBM. Upload it to FilmScribe as it is and export TXT. Our guide on how to transcribe video to text has the full walkthrough.
  • A YouTube video. Paste the link into FilmScribe, with nothing to download. Our guide on how to transcribe a YouTube video covers it, livestreams included.
  • A video that needs captions. Export SRT. YouTube lists SubRip (.srt) among the basic caption formats it supports and recommends it to people who are new to caption files, because it only needs the text and the timing. See how to create an SRT file.

What affects accuracy, in Word or anywhere else

No automatic transcript comes out perfect, and the W3C’s Web Accessibility Initiative is blunt about it: the text often doesn’t match the spoken audio, and missing a single word such as “not” can make it contradict what was actually said. Its advice is to plan time for corrections.

What matters most is the audio:

Audio problemWhat happens in the textHow to fix it
Wrong mic, or mic far from the mouthSwapped words, incomplete linesCheck the microphone input before you record
A call recorded with a headset onOnly your own voice shows upRecord without the headset, or record to a file
Echo (auditorium, church, empty room)Syllables drop or repeatRecord closer to the person speaking
Two people talking at onceMixed lines, wrong speakerListen to those passages again during review
Names and acronymsWrong spellingFix them in the text before it goes into the doc

During review, the WAI’s guidance for any transcript applies: identify the speakers, don’t change what was said, and mark a passage as unintelligible when you can’t make it out, rather than guessing.

Other options: OneNote, YouTube and your phone’s recorder

The same Transcribe exists in OneNote for Microsoft 365 on Windows, with the same formats and the same monthly cap; there, the ink notes you make while recording play back in sync with the audio. If the video is already on YouTube with captions, the Show transcript button gives you a rough draft to paste into Word, without speakers. And if the recording doesn’t exist yet, our guide on how to record and transcribe audio shows how to capture sound that turns into clean text. In all of these, Word comes in at the end, for formatting.

Common mistakes

  1. Playing the audio to Dictate. Dictation is for your own voice, live. An existing recording goes through Transcribe or an uploaded file.
  2. Closing the pane halfway. Word’s transcription needs the Transcribe pane open until it finishes.
  3. Spending the month’s minutes on silence. The intro, the break and the hallway chatter all count as uploaded minutes. Trim first, or mark just the range in FilmScribe.
  4. Starting a new transcript in the same document. The previous one disappears from the pane. Add what matters to the document first.
  5. Moving it into the document without review. One read with the audio alongside catches most errors in names, acronyms and numbers.

Frequently asked questions

Is Transcribe in Word free?

Not according to Microsoft's documentation: Transcribe needs an active Microsoft 365 subscription, and if you can't find the button, the first thing to check is that you're signed in with one. FilmScribe has no free plan either. Its trial, 3 days with 1 hour of material and a card, transcribes your real recording with speakers and timestamps.

What is the Word Transcribe limit?

Microsoft states a maximum of 300 minutes of uploaded audio per month on a Microsoft 365 subscription, which is 5 hours, and 30,000 minutes per month with a Microsoft Copilot license. The page lists no file size limit. Each document stores one transcript at a time.

What audio formats does Word Transcribe support?

Four, according to Microsoft: .wav, .mp4, .m4a and .mp3. WhatsApp voice messages in OGG or OPUS and video in MOV or MKV aren't on the list, so they need converting first. FilmScribe takes those files as they are, along with MP3, WAV, M4A, MP4 and YouTube links.

Can Word transcribe a video?

Only an .mp4 file, which is on Microsoft's list: upload it in the Transcribe pane. No other video format is listed, and the documentation describes no way to add a link. For MOV or MKV video, a long recording or a YouTube link, transcribe in FilmScribe, export TXT and open the text in Word.

Why is the Transcribe button missing in Word?

Check three things from Microsoft's documentation: that you're signed in with an active Microsoft 365 subscription; that, in Word for the web, you're on the new Microsoft Edge or Chrome; and that, in Word for Windows, your account belongs to a commercial tenant, the only kind the page says has the feature in the desktop app for now. Otherwise, try Word for the web.

Can Copilot transcribe audio to text in Word?

Microsoft's Transcribe page describes no separate Copilot workflow for audio files. What it says is that a Microsoft Copilot license raises the cap to 30,000 minutes of uploaded audio a month, against 300 minutes on a Microsoft 365 subscription alone. The steps are the same, Home > Dictate > Transcribe, with the same four file formats.

How do I turn a Word transcript into SRT captions?

Microsoft's Transcribe documentation only describes adding the text to the document, not a caption file. To get an SRT with in and out times for each line, transcribe the recording in FilmScribe and export SRT, with TV, Netflix and YouTube presets. The same file works on YouTube and in your video editor.

Sources

The official documentation behind the numbers and rules in this guide.

  1. 01Transcribe your recordings(opens in a new tab) · Microsoft
  2. 02Dictate your documents in Word(opens in a new tab) · Microsoft
  3. 03Supported subtitle and closed caption files(opens in a new tab) · YouTube Help
  4. 04Transcribing Audio to Text(opens in a new tab) · W3C WAI

Try it on your own recording. See the result before adding a card.

Upload a file or paste a YouTube link. Transcript, chapters, clips and captions are ready in minutes.