How to remove filler words (but know which to keep)

The word "um" being sawed in half by a giant saw

If you’re not a robot, you likely don’t speak in an uninterrupted, grammatically flawless manner.

When humans talk, it’s normal to subconsciously use short filler words — “um,” “uh,” “well,” “you know,” “I mean,” etc. — that fill brief pauses while we think of whatever we’re going to say next. These words also give others a verbal cue that we have more to say even if we’ve stopped speaking for a moment.

Filler words are perfectly natural in everyday conversation, but there are situations where it makes sense to remove filler words from podcast recordings. Keep reading for the lowdown on when you should and shouldn’t remove filler words from podcasts, and how to avoid the hassle of manually removing filler words by using Descript’s advanced AI tools.

Transcribe. Edit. As easy as tapping your backspace key.
Create your podcast from start to finish with Descript.

Why remove filler words?

Here are a few of the main reasons why you’d consider removing filler words from your podcast:

  • To clean up a scripted podcast. Filler words fit in naturally in podcasts centered around casual conversations. But it makes sense to remove filler words from more formal scripted podcasts like narrative audio dramas, storytelling-focused shows, and investigative news reports — or really any show with a more polished tone.
  • To shorten an episode. One filler word may only take a millisecond or two of time, but that time adds up quickly when you consider all the filler words that can slip in over the course of an episode. If your episode feels slow-paced or you’re slightly over your target length, cutting filler words can help shorten it without losing valuable content.
  • To avoid annoying listeners. If you’ve ever listened to a podcast and thought, “Can you please just get to the point already?” or found yourself waiting for the next time the host mutters “um,” you know what we’re saying here. Excessive filler words can quickly annoy listeners, so removing them might help you hold their attention.

What to consider when removing filler words

Before you go on a filler-word removing frenzy, here are a few things to think about.

  • Removing too many filler words can sound unnatural. It’s fine to remove a few "um" sounds here and there, but avoid going overboard. Editing out every single filler word tends to make a speaker sound robotic and inauthentic.
  • Filler words belong in certain show formats. If you remove filler words from conversational types of shows, you risk making the show feel too formal and less engaging. And think twice if you’re making a comedy podcast — removing filler words risks ruining the comedic timing of certain jokes. But if your show is scripted, you probably don't want filler words; same in voice over.
  • Ask yourself if it’s worth your time. Even with Descript’s easy filler word removal tool, removing filler words adds time to the editing process. You still have to choose which filler words to remove and listen to the audio afterward to ensure it sounds correct. So avoid removing filler words just for the sake of it and make sure your podcast or episode calls for it.

How to remove filler words from audio using Descript

Descript’s convenient filler word removal feature instantly detects superfluous filler words and removes them from your audio track. Here’s how to get rid of filler words with just a few clicks:

Import and transcribe your audio file. Create a new project and either drag and drop your audio file into Descript or click the “Choose a file…” button to import it. Descript will then begin to automatically transcribe your file. The transcription of a 60-minute podcast typically takes around two to three minutes.

Look over the detected filler words. Once Descript completes your podcast transcription, it instantly labels filler words within the transcribed text using a light-blue underline. By default, Descript detects the following filler words and phrases:

  • “but you know”
  • “hm”
  • “I guess”
  • “I mean”
  • “I suppose”
  • “kind of”
  • “like”
  • “mmm”
  • “or something”
  • “right”
  • “so”
  • “sort of”
  • “uh”
  • “um”
  • “well”
  • “you know”
  • “you know what I mean”
  • “you see”

Delete filler words manually using the transcript. There are a couple different options for removing detected filler words. To remove words one-by-one in your transcription, simply highlight a filler word with your cursor and hit the “delete” key. Descript then deletes that word from your audio track and automatically crossfades the newly conjoined regions, creating a pretty seamless edit. You can identify them yourself, or run Descript's Filler word detection, which will underline each filler word in light blue.

Delete filler words automatically using the “Remove Filler Words” tool. Click the wrench icon at the top of your transcription and select “Remove filler words…” This brings up a sidebar displaying every filler word the app detected, and giving you two removal options for each: “Delete” or “Ignore.” The “Delete” button removes the filler word from both the audio track and the transcription. The “Ignore” button removes the word from the audio track, but only crosses out the word in the transcription, so you have a visual record of its being there. You can also select “Replace with gap clip” to add a gap in the audio equal to the length of the spoken word. To apply your changes for each word, click the “Apply” button. To remove every filler word with one click, select “Apply to all.”

PRO TIP: Need to restore a removed filler word? Go to “Edit” > “Undo” or press Command + Z (MacOS) or Control + Z (Windows). If you used the “Ignore” function, you can also restore the word at any point by highlighting it within the transcript and clicking the “Strikethrough” button in the top toolbar.

How to remove filler words from video using Descript

Podcasters often film their podcast recordings to release as full video podcasts, or to edit short promotional video clips for use on social media. Since Descript is both an audio and video editor, you can use its filler word removal tool on video podcasts too.

To remove filler words from video, follow the same steps for removing filler words from audio (above) but pay extra attention to the “replace with gap clip” option. Here’s how turning the gap clip option on or off will affect your video:

  • Toggling "Replace with gap clip" ON: This removes filler words from the audio, but the video remains unedited. Selecting this option essentially mutes the filler words from the video. It avoids jump cuts but can create awkward moments where you see the speaker's lips moving but don't hear their voice.
  • Toggling "Replace with gap clip" OFF: This removes the audio and video associated with the filler words but creates a noticeable jump cuts.

Featured articles:


Here's how to turn a mountain of interview tape into a coherent podcast episode

It can feel overwhelming to take everything you’ve recorded and try to pull it into a single, coherent podcast episode. We asked an expert how to sort through it all.


Fine cuts: Making your video screen-ready

A fine cut is where you sprinkle on the finishing details and add things like music, graphics, color correction, and whatever adjustments and corrections your video needs to be ready to publish.


YouTube transitions that will level up your video

Almost every piece of media we consume has some kind of transition, whether we notice it or not — and not noticing is usually the goal.

Related articles:

Share this article

See also