Microsoft is adding an audio transcription feature to the online version of Word that will be free for Office 365 subscribers.

The new service will allow users to import existing audio files or to record conversations directly into Word before having them automatically transcribed.

Microsoft’s transcription feature can capture audio from your PC, which can be MP3 recordings of meetings, phone calls or even YouTube videos. It will also capture audio from your machine’s microphone for direct dictation. It has support for up to 200MB of MP3, WAV, M4A or MP4 files.

Once a conversation is transcribed, Microsoft’s AI will separate different speakers and break the text into easily readable chunks that will be timestamped for easier playback. Users can then edit and insert these text clippings into a Word document.

The playback element is similar to its text-to-audio service, Immersive Reader, which was launched on 25 August and powered by its Azure AI platform.

AI-powered transcription has grown in popularity in recent years, particularly with apps like Otter.ai, Happy Scribe and Trint, all of which boast high degrees of accuracy when transcribing near-perfect audio, but fall short when it comes to non-US accents.

However, it remains to be seen whether Microsoft’s own software is going to be able to draw users away from existing free models, particularly as there’s no way to access it outside of Office 365 and usage is capped at 5 hours worth of uploads per month, which is half the upload allowance available on the free tier of Otter.ai.

Transcribe in Word will be available from today and is free for all Microsoft 365 subscribers. It’s also supported in the new Microsoft Edge and Chrome browsers. A mobile version for iOS and Android is due later in the year.

The service currently only supports English, although Microsoft said more languages will be made available at a later date.

Gif courtesy of Microsoft Blog