What is Descript?
Descript is an audio and video editor with a genuinely clever twist: instead of working with a timeline full of clips and waveforms, you edit a text transcript. When Descript transcribes your recording, every word is linked to the exact moment it was spoken. Delete a sentence in the transcript and that audio or video simply disappears. Move a paragraph and the recording rearranges itself to match.
For podcasters, YouTubers, and anyone who records meetings or presentations, this completely changes the editing experience. You don't need to scrub through waveforms hunting for the bit you want to cut — you just read the transcript and edit it like a document.
Descript also includes a feature called Overdub, which clones your voice. Once trained on a sample of your speech, it can generate new audio in your voice from typed text. This means you can fix a stumbled sentence in a podcast episode without re-recording it.
In 2026, Descript has expanded significantly with Underlord, an AI co-editor that can write, edit, design, and generate visuals and voiceovers. The platform now also offers AI avatars, AI background removal, eye-contact correction, and advanced noise removal — making it a comprehensive video production tool, not just an editor.
Best Use Cases
Podcast Editing
Upload your raw recording and let Descript transcribe it automatically. Then read through the transcript, highlight any rambling sections or mistakes, and delete them. You can remove all filler words like "um" and "uh" in a single click using the built-in filler word removal tool.
Try this prompt (Overdub): Type the corrected version of a stumbled sentence directly into the transcript and Overdub will regenerate just that line in your voice — seamlessly replacing the mistake.
Video Editing
Record a tutorial, interview, or explainer video and import it into Descript. Use the transcript to cut out long pauses, false starts, and off-topic tangents. You can also add captions automatically from the same transcript, saving hours of manual work.
Try this prompt: After importing your video, select all instances of a repeated phrase you want to remove, right-click, and choose "Delete all occurrences" — Descript removes every instance from the recording at once.
Screen Recording
Use Descript's built-in screen recorder to capture tutorials or walkthroughs. The recording is immediately transcribed and editable, so you can trim your hesitations and tighten up the narration before publishing.
Try this: Record a short software tutorial, then use the transcript to cut any section where you navigated to the wrong menu or had to redo a step — making the final video look clean and intentional.
Getting Started
- Go to descript.com and create a free account (the free plan includes 1 hour of transcription per month).
- Click New Project and choose whether you're working with audio or video.
- Import your recording — Descript will automatically transcribe it within a minute or two.
- Read through the transcript and select any text you want to remove. Press Delete and it will vanish from both the transcript and the recording.
- When you're happy with your edit, click Publish or Export to download your finished file.
Tips for Beginners
- Use "Remove filler words" first. Before doing any manual editing, run the automatic filler word removal. It instantly cleans up a huge portion of the rough edges in most recordings.
- Edit in chunks, not word by word. Select whole sentences or paragraphs to cut — trying to trim individual words often sounds unnatural because you lose the breathing rhythm.
- Train Overdub on at least 10 minutes of clean speech. The more varied the sample, the more natural-sounding the cloned voice will be.
- Keep your original file. Descript's edits are non-destructive by default, but always keep a backup of your raw recording before you publish.
- Use Scene view for video. Switch to Scene view when working with video to see your footage divided into visual segments that match the transcript paragraphs — it makes restructuring much easier.