Extracting audio in CapCut — and actually getting the file

CapCut separates audio in one tap. Getting that audio out as a file is a different question, and the answer differs between the phone and the desktop app.

CapCut separates audio from video in one tap. Getting that audio out as a file you can use somewhere else is the part that stumps people, and the answer is different on the phone and on the desktop.

On the phone: extract, then export the workaround

In the mobile app, adding a clip and tapping Audio → Extracted — or long-pressing the clip and choosing Extract audio — puts the soundtrack on its own track below the video. That part is easy and works exactly as advertised.

What the mobile app does not offer is an audio-only export. Every Export produces a video file. The usual workaround is to delete the video track, leaving a black screen with the audio, export that as a video, and then pull the audio out of the export — which is a lot of work to arrive back where you started.

If the goal is an audio file rather than an edit, skip the editor: drop the original video here and take the soundtrack out directly. On a phone, Chrome on Android runs this page fine; on an iPhone, these four routes cover it.

On the desktop: there is a real audio export

CapCut for Windows and Mac does have one. Put the clip on the timeline, extract the audio if you want it separated, then in the Export dialog switch off the video option or choose the audio export tab, where MP3 and WAV are offered. That is a genuine audio-only render and it is the right route if you are editing anyway.

Why an editor is the wrong tool for a straight extraction

Editors re-encode. CapCut imports your video, decodes it, renders a timeline and encodes the result — which takes real time, applies its own quality settings, and can add a watermark depending on the template and the version you are on. For an actual edit that is the price of doing business. For "I just want the sound", it is all cost and no benefit.

Taking the soundtrack out of a container does not require decoding anything at all. When the format allows it, the audio is copied out exactly as it was recorded — seconds, identical, and no settings to get wrong. The explanation, with the measurements.

The watermark question

CapCut's watermark applies to the picture. An audio-only export from the desktop app has nowhere to put one, so the audio comes out clean. If you used the "delete the video track and export a black video" workaround, whatever the app decided to overlay is in the video, and the audio you extract from it afterwards is unaffected.

If you do export from CapCut

• Choose MP3 at the highest bitrate offered if the file is going elsewhere, or WAV if it is going into another editor.
• Remember that this is a second encode of audio that was already compressed in the original video. It is not ruinous, but it is not free.
• Check the length: exporting a timeline exports the timeline, including any trailing silence you left at the end.

The two-step that most people actually want

Extract first, edit second. Take the audio out of the video with something that does not re-encode it, then do whatever trimming or fading you need on the audio file itself — which is faster, keeps the quality, and does not involve a timeline at all. Our sister site Audio Converter handles the audio-editing half; this page handles the first step.