How to Use Descript: A No-BS Guide to Text-Based Video Editing

Descript is one of those tools that actually delivers on its promise: edit video and audio by editing text. No timeline scrubbing. No frame-by-frame adjustments. Just delete words from a transcript and the corresponding video disappears. It's genuinely different from traditional video editors like Premiere or Final Cut.

If you're here because you just downloaded Descript and have no idea where to start, I've got you covered. This guide walks through everything from your first project to exporting a polished video.

Already know about Descript's features but need pricing info? Check out our Descript pricing breakdown or full Descript review.

What Descript Actually Does

Descript combines audio editing, video editing, screen recording, and transcription into one platform. The key differentiator is text-based editing-you edit your media by editing the transcript, which automatically updates the corresponding audio and video.

The AI-powered features automate tedious tasks like generating transcripts, removing filler words, and even identifying different speakers automatically. Advanced functions like multi-track recordings and speaker identification happen in the background, so you don't need to know a lot to get started.

Here's what you can do with it:

Getting Started: Your First Descript Project

Step 1: Download and Sign Up

Head to Descript's website and download the desktop app. It's available for both Mac and Windows. Note that there's no mobile app-Descript is desktop-only.

The free plan gives you 60 media minutes per month and 100 one-time AI credits. It's enough to test the waters, but you'll burn through those limits quickly just playing around. The free plan exports are watermarked once per month for watermark-free exports, and video resolution is capped at 720p.

Step 2: Create a New Project

When you open Descript, you'll land in the Drive View-think of it like your file manager for all your projects. From here, you can see all your projects, quick recordings, and manage your workspaces.

To start a new project:

Step 3: Import or Record Your Media

You have two options: import existing files or record directly in Descript.

To import: Drag and drop your file into the Script editor, or click the + icon and select File. Descript supports various file formats including MP3, WAV, MP4, MOV, Apple ProRes, and QuickTime RLE. The maximum file length for transcription is 15 hours-split longer files into smaller segments before uploading to prevent performance issues.

To record: Click the Record button to open the Record panel. You can record audio only, video (webcam), or screen recordings. Descript even has a built-in teleprompter you can use while recording.

Once your file is imported, Descript automatically starts transcription. This takes a few minutes depending on file length. You can also choose to transcribe during recording by enabling the "Transcribe" option in Additional settings next to your audio input.

Understanding Descript's Interface

The Drive View

The Drive View is your home base in Descript. It has three key sections:

1. Project Panel: Create new projects or jump back into existing ones. You can move, rename, or delete projects and adjust project access to keep everything organized. Use folders to organize your projects and keep your team's workspace tidy.

2. The Sidebar: Switch between drives, purchase new drives, invite team members, and access different tabs:

3. Project Files Section: Once you're inside a project, the Project panel shows all your media files, compositions, and sequences.

The Composition Editor

When you create or open a composition, you'll see the main editing interface with several key areas:

The Basics of Text-Based Editing

This is where Descript gets interesting. Once your transcript is generated, you edit video like you'd edit a document.

Deleting Content

See a section where you rambled? Highlight the text and hit delete. The corresponding audio and video disappears. It's that simple.

Cutting Scenes

For more precise control, use the split tool. Position the playhead where you want to cut, then press Command+E (Mac) or Ctrl+E (Windows). Click on unwanted segments and delete them.

Rearranging Content

Cut and paste text to reorder your video. The audio and video follow the transcript. This makes rearranging interview answers or podcast segments incredibly fast.

Correcting the Transcript

The AI transcription is good but not perfect-accuracy reaches up to 95% with clear audio, but varies with recording quality, accents, background noise, and mic placement. Click on any word to correct mistakes, or press 'C' to enter correction mode. This doesn't change your audio-it just fixes the text for captions or reference.

Working with Transcription

Supported Languages

Descript can automatically transcribe audio and video in 26 languages, all using the Latin alphabet. Supported languages include:

Languages like Chinese, Japanese, or Russian are not yet supported, as Descript currently only supports Latin alphabet languages.

Setting Your Transcription Language

If you work with multiple languages, set a preferred transcription language in your App settings or enable "Always ask before transcribing." When enabled, Descript will prompt you to select a language before each transcription.

If you transcribed a file in the wrong language, you can re-transcribe it: Click Project in the right-hand sidebar, open the Files section, click the menu next to your file name, select Re-transcribe, and choose the correct language.

Transcription Limitations

Keep in mind that Descript only supports one language per file-files with multiple spoken languages aren't currently supported. You'll need to create separate compositions for different sections if your content switches between languages.

Also note that Descript transcribes spoken content but won't produce song lyrics or musical notation. Music-heavy parts typically show up as blank or untranscribed.

Improving Transcription Accuracy

For the best transcription results:

Using the Timeline

Text-based editing handles most tasks, but sometimes you need the timeline for finer control.

The timeline shows your audio waveforms and video tracks. You can:

The Properties panel on the right side lets you adjust everything in a scene or layer-position, size, effects, and more.

Advanced Multitrack Editing with Sequences

What Are Sequences?

A sequence in Descript is a virtual container that groups multiple media tracks together while maintaining their time alignment. Think of sequences as the glue that binds separate recordings together, ensuring they stay synchronized throughout the editing process.

Sequences are automatically created when you record in Descript Rooms or use the Editor recorder. They're essential for:

Creating a Sequence

There are several ways to create sequences:

Method 1: Import Multiple Files

  1. Drag multiple files into your composition at the same time
  2. When prompted, select "Combine into sequence"
  3. Assign speaker labels to each file
  4. Descript creates a sequence that keeps all tracks aligned

Method 2: From Existing Files

  1. Select files in the Project panel
  2. Right-click and choose "Create sequence"
  3. Descript bundles them and opens the Sequence Editor for alignment adjustments

Method 3: Start Empty

  1. Click the three-dot menu in the Project panel under Files
  2. Select "Create sequence"
  3. Manually add files and build your sequence in the Sequence Editor

Editing Sequences

The beauty of sequences is that you can edit all tracks at once from your composition's script. When you delete text or rearrange sections, all tracks in the sequence update together automatically.

For individual track adjustments, open the Sequence Editor by right-clicking the media and selecting "Edit sequence." This lets you:

Descript supports up to 14 sequence tracks, providing enough room for complex multi-camera or multi-mic setups.

Multitrack Audio Mixing

When working with sequences, you have precise control over each audio track:

For podcasts with multiple speakers on separate tracks, multitrack transcription provides perfect speaker labeling precision and more accurate transcripts by isolating voices.

Descript's AI Features (The Good Stuff)

Descript's AI tools are accessible through the "Underlord" button in the top right corner. These features are what make Descript worth using over traditional editors.

Studio Sound

This is Descript's professional-grade audio restoration. It removes background noise, room echo, and enhances voice quality with one click. The intensity slider goes from 0-100, but pro tip: leaving it at 100% can make audio sound unnatural with weird artifacts. Start around 75% and adjust from there.

Remove Filler Words

Automatically removes "um," "uh," "like," repeated words, and other verbal crutches. You can review what it wants to remove before committing. This feature saves hours of manual editing on podcast and interview content.

Shorten Word Gaps

Automatically removes silences and dead air to tighten up your pacing. You can adjust the threshold to control how aggressive the silence removal is.

Eye Contact

This AI effect adjusts your gaze so it appears you're looking directly at the camera, even if you were reading from a script. It works best with single-person videos where your face is clearly visible and well-lit. If you wear glasses, softer angled lighting helps reduce glare.

Eye Contact is non-destructive-toggle it on or off anytime. Note: it only works with one person in frame. Group videos won't apply the effect.

AI Green Screen

Remove or replace your video background without an actual green screen. Enable it and swap your background with an image or stock clip. Great for hiding messy offices or creating consistent branded backgrounds.

Edit for Clarity

The AI suggests parts of your content to remove for better flow. It analyzes your transcript and identifies redundant sections, long pauses, or content that doesn't add value.

Underlord Chat (AI Co-Editor)

This is Descript's AI co-editor. You can chat with it to repurpose content, refine edits, or generate new content. It uses GPT-4o under the hood, so if you're familiar with prompting chatbots, you'll feel right at home.

Use Underlord to:

AI Credits System

As of September 2025, Descript uses a credit system for AI features. Different AI tools consume different amounts of credits:

Unused AI credits do not roll over month-to-month. The Free plan includes 100 one-time AI credits, while the Creator plan includes 800 AI credits per month.

Using Overdub: AI Voice Cloning

What Is Overdub?

Overdub is Descript's text-to-speech AI voice generator that lets you create a realistic clone of your own voice. Once you've created a voice clone, you can generate speech just by typing-perfect for fixing mistakes, adding content, or creating voiceovers without re-recording.

Creating Your Voice Clone

To create a custom Overdub voice:

  1. Click on "Overdub" in the Drive view
  2. Record yourself reading a provided script (minimum 10 minutes, but 30 minutes recommended for best results)
  3. Record in a quiet, acoustically "dead" room using an external microphone
  4. Descript processes your recording and generates your custom voice
  5. Fine-tune voice settings like pitch and speed to achieve the desired output

You can create multiple Overdub voices to fit different performance styles or recording settings. For example, create separate voices for Zoom recordings versus studio recordings.

Voice clones on Free and Creator accounts have vocabularies of 1,000 common words. Pro accounts get unlimited vocabularies. Descript requires explicit consent to create voice clones-you can only clone your own voice, not someone else's.

Using Overdub in Your Projects

Once your voice is ready:

  1. Type the text you want to convert to speech in your script
  2. Click on the speaker label and select your Overdub voice
  3. The AI generates audio that sounds like you speaking
  4. Make edits to the text and the audio updates automatically

Stock AI Voices

If you don't want to clone your voice, Descript offers a library of ultra-realistic stock AI voices. These voices speak over 20 languages with a full range of emotions. Popular stock voices include Cedric, Carla, and Emily.

Stock AI voices don't just pause at commas and add inflection at question marks-they vary tone and rhythm naturally as they speak, making them sound surprisingly lifelike.

Regenerate Feature

The Regenerate feature uses AI to smooth out awkward edits and fix mistakes. When you make a cut that sounds abrupt, Regenerate can generate new audio that blends seamlessly with the surrounding content. It's like magic for fixing obvious cuts or balancing noise levels.

To use Regenerate:

  1. Highlight the section you want to update
  2. Click Regenerate in the selection toolbar or press D
  3. Descript generates new audio matching your voice

Note: Regenerate requires a custom AI speaker and doesn't support stock voices.

Tips for Better Overdub Quality

Adding Captions, B-Roll, and Other Elements

Captions

Descript automatically generates captions from your script. Open the Captions panel in the sidebar, choose a style, and customize colors, fonts, and positioning to match your video. The "Fancy Captions" feature creates animated captions that are perfect for social media content.

You can export captions in various formats including SRT, VTT, plain text, rich text, markdown, HTML, or Word doc. Captions can also be translated into 30+ languages using Descript's AI translation feature.

Stock Media

Need b-roll footage, music, or sound effects? Descript has a built-in stock media library. Add shapes, text, images, and other visual elements directly from the editor. The library includes hundreds of music tracks and sound effects.

Templates

Start with pre-built templates for common formats-YouTube videos, social clips, podcasts-or save your own layouts for consistency across projects. Templates are especially useful for teams that need to maintain brand consistency.

On paid plans, you can create custom templates that include your brand colors, fonts, logos, and layouts. This centralizes your brand elements so every team member can quickly create on-brand content.

Screen Recording in Descript

Descript includes a standalone screen recorder for capturing tutorials, demos, or presentations. You can record your screen, webcam, or both simultaneously.

Recording Options

The screen recorder offers several capture modes:

On Mac, you can resize your webcam bubble and choose whether it's a square or circle. The webcam overlay never appears in your final recording-it's only visible to you during capture.

Built-in Teleprompter

The built-in teleprompter displays your script while recording. It scrolls automatically as you speak (or manually if you prefer), helping you stay on message without looking away from the camera. The teleprompter window never appears in your final recording.

Recording Best Practices

For more options, check out our guides to free screen recording software and best screen recording tools.

Collaboration Features for Teams

Drive Membership Roles

Descript has two types of Drive membership roles:

Editor (paid seat): Full access to the Drive workspace. Editors can create, edit, move, and delete projects and folders. They use Media Minutes when they record or import media, and AI Credits when they apply AI-powered effects.

Viewer (free seat): Can view and comment on projects in the Drive. Viewers cannot edit content or use features that consume Media Minutes or AI Credits.

The Drive owner is the person who created the Drive and has full Editor permissions plus administrative privileges.

Project-Level Collaboration

You can invite collaborators to a specific project without adding them to your entire Drive. This is ideal for one-off collaborations, collecting feedback, or giving someone limited access without granting full Drive access.

Project collaborators come in two types:

To invite someone to a project, click the sharing icon in the top right corner of your project, enter their email, and select their permission level. They'll receive an email invitation and can access the project with a free Descript account.

Real-Time Collaboration

Multiple users can work on the same project simultaneously. When you're working in a project at the same time as teammates, you'll see their avatars and can watch their edits happen in real-time-similar to Google Docs.

Team members can:

Version History

Nothing is permanent in Descript. The platform automatically saves changes as you work, and you can easily revert to earlier versions if you make a mistake. Since edits are non-destructive, source files remain intact.

Team Workspace Organization

Organize your work with folders and automatically share projects with your team. Features include:

Integration with Other Tools

Descript works seamlessly with tools your team already uses:

Exporting and Sharing Your Project

When you're ready to share, Descript gives you several options:

Local Export

Save to your device in formats like MP4, MP3, WAV, or GIF. Export resolution depends on your plan:

You can also export audio into discrete chunks divided by line breaks or markers, or export every composition in a project in one go.

Publish Link

Generate a hosted version of your project with a shareable URL. Viewers can leave timestamped comments directly on the video-useful for collecting feedback without switching tools.

Publish settings let you customize:

The publish option uses Descript's servers to render your video, which is the fastest way to share. Click Publish in the top right, configure your settings, and you'll get a URL to send around.

Platform Integrations

Export directly to YouTube, HubSpot, and other platforms without downloading first. This streamlines your workflow and saves time on file transfers.

Export for Professional Editors

If you're using Descript for rough cuts before final editing in Premiere or Final Cut, you can export FCPXML files or AAF sequences that maintain your edits and transitions. This is useful for teams that need advanced color grading, effects, or motion graphics work.

Descript Pricing: What You Actually Pay

Descript has multiple pricing tiers. Here's what they cost (annual billing):

The free plan is honestly pretty limiting-you'll hit the 60 media minute cap quickly just testing features. If you're serious about using Descript, plan on at least the Hobbyist tier.

Educational institutions and non-profits can access a special discounted plan for $5 per user per month, which includes the same features as the Creator plan but with a 4-hour monthly transcription limit.

Understanding Media Minutes and AI Credits

As of September 2025, Descript moved from transcription-hour limits to a system based on Media Minutes and AI Credits:

Media Minutes: Track uploaded media files and media recorded in Descript, regardless of whether they're transcribed. Still images count as 1 second each.

AI Credits: Track your usage of AI features like Underlord, Studio Sound, Green Screen, Eye Contact, and AI-generated media and avatars.

Both reset monthly with no rollover. Descript says a la carte top-ups are "coming soon" for when you exceed your plan limits.

For a full breakdown, see our Descript pricing guide.

Tips for Getting the Most Out of Descript

Use Keyboard Shortcuts

Press ⌘K (Mac) or Ctrl+K (Windows) to open the Search Actions feature. Type what you want to do, and Descript will help you find the right feature or execute commands. This is faster than hunting through menus.

Master the Essential Shortcuts

Version History is Your Friend

Don't be afraid to experiment. Use Version History to restore earlier versions of your project if you make a mistake. This safety net means you can try bold edits without fear.

Don't Over-Process Audio

Studio Sound at 100% can sound artificial. Start at 75% and adjust based on your source audio quality. Less is often more when it comes to AI audio enhancement.

Collaborate Like Google Docs

Share projects with teammates who can view, comment, or edit based on permissions you set. Projects sync to the cloud automatically-you can even share with people who don't have Descript accounts for viewing and commenting.

Use the Transcription Glossary

Add industry jargon, technical terms, or proper names to your Transcription Glossary so Descript transcribes them correctly every time. This is especially useful for podcasts or videos with specialized vocabulary.

Create Templates for Consistent Branding

If you're making multiple videos with the same format, create a template with your intro, outro, caption styles, and brand colors. This saves time and ensures consistency across all your content.

Optimize Your Recording Setup

Common Workflows in Descript

Podcast Editing Workflow

  1. Import your multitrack recording (or record directly in Descript Rooms)
  2. Let Descript transcribe and identify speakers
  3. Remove filler words with one click
  4. Delete rambling sections by editing the transcript
  5. Apply Studio Sound to clean up audio
  6. Add intro/outro music and any sound effects
  7. Generate show notes and social clips with Underlord
  8. Export as MP3 for your podcast host

YouTube Video Workflow

  1. Import your video or use screen recording
  2. Edit out mistakes and dead air from the transcript
  3. Add Eye Contact effect if needed
  4. Shorten word gaps to tighten pacing
  5. Add captions for accessibility
  6. Insert b-roll over sections where you're describing something visual
  7. Create chapter markers
  8. Export or publish directly to YouTube

Social Media Clips Workflow

  1. Import your long-form content
  2. Use Underlord to identify the best clips
  3. Resize for vertical format (9:16)
  4. Add Fancy Captions for engagement
  5. Apply AI Green Screen for clean backgrounds
  6. Export multiple clips as separate files

Interview Editing Workflow

  1. Import your interview recording
  2. Review the transcript and identify key moments
  3. Rearrange answers to improve narrative flow
  4. Remove verbal tics and long pauses
  5. Use Edit for Clarity suggestions
  6. Add speaker labels and ensure they're accurate
  7. Export the finished interview

What Descript Doesn't Do Well

Let's be honest about the limitations:

Descript vs. Traditional Video Editors

When Descript Wins

When Traditional Editors Win

The Hybrid Approach

Many professionals use Descript for rough cuts and initial editing, then export to Premiere Pro or Final Cut for final polish. This combines the speed of text-based editing with the power of professional NLEs.

Security and Privacy

Descript takes security seriously:

For Enterprise customers, additional security features include Single Sign-On (SSO), custom security reviews, and compliance with organizational security requirements.

Customer Support and Resources

Descript provides several support channels:

The in-app help feature (click the ? in the bottom right) provides context-sensitive help and quick answers without leaving your project.

Bottom Line: Should You Use Descript?

If you're making podcasts, YouTube videos, social content, or screen recordings, Descript is genuinely useful. The text-based editing approach is faster than traditional timeline editing for most spoken-word content, and the AI features (especially Studio Sound and filler word removal) save real time.

The learning curve is much gentler than professional editors like Premiere or Final Cut. If you can edit a Google Doc, you can edit in Descript. This makes it accessible for content creators, marketers, educators, and business professionals who need to create video content without becoming video editing experts.

For teams, the collaboration features make Descript a solid choice. Real-time editing, version history, and project-level sharing remove friction from collaborative workflows. The ability to invite viewers for free means you can collect feedback from stakeholders without paying for extra seats.

The AI features-voice cloning, automatic transcription, translation, Studio Sound-are legitimately impressive and continue to improve. Underlord's ability to repurpose content into blog posts, social clips, and summaries adds real value for content creators managing multiple channels.

However, the pricing structure with media minutes and AI credits can feel restrictive, especially for heavy users. The lack of rollover is frustrating if your usage varies month to month. And if you need advanced color grading, complex effects, or frame-level precision, traditional editors remain the better choice.

For complex visual projects with lots of effects, motion graphics, or multi-camera setups with frequent angle switching, you'll probably want a traditional editor. But for everything else? Descript is worth trying.

Try Descript Free →

Looking for alternatives? Check out our guides to free video editing software and best video editing software for more options.