Apps I use (and you should too): Descript

Quick transcription, editing, and recording

Apps I Use is a series about software that I rely on. It’s not sponsored or an ad—it’s just apps I enjoy and that you might find useful, too.

When I left my job at Digital Trends in 2020, I suddenly needed a whole suite of productivity apps to replace the employer-mandated apps I’d used prior.

One of the first productivity challenges I needed to solve was specific but crucial: transcription.

I spend a lot of time talking to people. Often, those conversations need to be recorded and converted into a transcript I can browse and search.

So, I went looking and found Descript. It’s an app for audio transcription—and more.

What is Descript?

Descript launched in 2017 as an audio editing app with a strong focus on transcription. Initially, users could upload audio files and have them transcribed at a per-minute rate.

Since then, the app has broadened its scope significantly. It now markets itself as an AI-powered video editor that “lets you edit videos and podcasts like a doc.” I'm not sure this is the best marketing—it’s a bit confusing, isn’t it?

Basically, Descript supports a wide range of audio and video files, which are transcribed (with AI) into a document.

The app can also record audio and video directly, and even has an AI voice clone feature that lets you add new spoken audio (in your own voice, if you’d like) by adding text to the document.

How do I use Descript?

My use of Descript is focused on its original mission: transcription.

When I finish a recorded interview with a source, I immediately upload that recording to Descript. The app transcribes the file (usually in two or three minutes). I also run speaker detection to add names to each speaker.

As an added bonus, this provides a cloud-based backup for my file. I’ve yet to lose a locally recorded audio or video clip, but it’s wise to have a backup.

I also use a recently added feature, called Underlord (again, I’m not sure about the company’s marketing…), which uses AI to shorten and clarify the transcript.

Underlord removes long pauses, filler words (the “ums” and “ahs”), and can edit down the transcript for clarity.

The result is a shorter, more palatable transcript, though the original audio remains accessible if it’s needed.

What I like about Descript

At its core, Descript is built around using AI to convert audio to text. It does that with stunning speed and accuracy.

This app can ingest an hour of dialogue and audio and spit out a transcript in a few minutes.

And while I find it confusing, I must admit there’s something to Descript’s whole “edit video and podcasts like a doc” approach.

Descript opens audio transcripts as a document I can edit.

While I don’t spend a ton of time editing the transcript text in Descript, I spend a lot of time using the document editor’s markup features to highlight sections of text I want to remember or quote while writing an article.

Here’s the secret sauce: with Descript, the audio remains attached to the document.

If I’m unclear about a portion of the transcript, or I’m cutting a quote to fit in my word count, I can Alt-Click the text I want to hear. Descript will pick up from where I clicked.

Descript also has a great search function. Sections that match my query appear in a sidebar. I can click any result to immediately hear that section of audio.

What I don’t like about Descript

Opening Descript for the first time can leave users a bit puzzled, because there’s a lot going on. It’s trying to be a transcription app, podcast editor, and audio/video recorder in one.

You might ask: why does a transcription app include options for visual waveforms and progress bars? Why does a podcast editor have a toolbar for adding animated gifs? Why does a video editor include text markup features?

It began to make sense as I learned to use the app, but it took a minute.

Descript won’t run on a potato, either. The text interface makes it seem like an app that should run on any old rig but, if you’re working with video files, it can require a lot of RAM.

It brings my poor Mac Mini with M1 chip and 8GB of RAM to its knees. The entire system slows to a crawl when I transcribe and view a video file.

A Surface Laptop with Snapdragon X Elite and 32GB of RAM doesn’t have a problem, however.

How much does Descript cost?

While I call Descript an “app,” it’s technically a service. You need to subscribe to use it, and the app is useless without an Internet connection.

Pricing starts at $12/month for the Basic plan, which supports 10 transcription hours a month and 1080p video editing. Many remote workers will need to upgrade to the $24/month Creator plan, as that allows 30 transcription hours a month and supports 4K video.

Will Descript work on my device?

Descript has an app available for both Windows and MacOS.

The website also provides a web app that works in any modern web browser. That makes it accessible on Chromebooks, and even on an iPad or Android tablet (though I can’t comment on how well the UI holds up on a touchscreen).

So, should you use Descript?

If you need transcription, yes, absolutely. While far from the only way to transcribe audio these days, Descript’s edit-audio-as-a-document interface and quick search function are tough to beat.

Also give it a look if you want a simple way to record podcasts and share them as both audio and text.

Want to use Descript? Get it here.