Extract Audio from Video

How to extract audio from a video file

Extracting audio from a video means separating the audio track from the visual track and saving it as a standalone audio file. MP4 to MP3 does this with FFmpeg, the open-source media framework trusted by virtually every professional video pipeline. When you upload a video, FFmpeg demuxes the container, isolates the audio stream, re-encodes it as MP3 using the libmp3lame encoder, and discards the video entirely — leaving you with a clean, portable audio file.

The whole extraction takes just a few seconds for typical videos under 100 MB. There is no queue, no email verification, no waiting for a download link to arrive in your inbox. Click Extract Audio, wait for the spinner to finish, and your MP3 is delivered straight to your browser as a downloadable file.

This tool is built for your own videos — files you recorded, exported from your editor, or have explicit permission to process. It does not fetch, scrape, or download content from Instagram, TikTok, YouTube, Facebook, or any other third-party platform. If you are looking for a way to download audio from someone else's Reel or video, please use the official share features on those platforms instead — scraping copyrighted content is illegal and against our Terms of Service.

Why extract the audio track separately?

There are many legitimate reasons to separate audio from video. Podcasters often record interviews on Zoom or Riverside and need just the audio for their editor. Voiceover artists record scratch takes on their phone and want a portable MP3 for the studio. Students record lectures and prefer to listen on the commute. Translators and transcriptionists work faster from audio than video. Whatever your reason, isolating the audio gives you a smaller, more portable, more editable file.

Because the output is a standard MP3, the resulting file plays on every modern device and audio app — from iOS Music, Android players, and VLC to Audacity, Premiere Pro, GarageBand, and Discord uploads. There is no DRM, no special codec requirement, and no container to worry about. The MP3 will still play perfectly in 20 years.

Technical details of the extraction process

We use FFmpeg with the following command structure: ffmpeg -i input -vn -map a -acodec libmp3lame -q:a 2 output.mp3. The -vn flag tells FFmpeg to ignore the video stream, -map a selects only audio streams, libmp3lame is the MP3 encoder, and -q:a 2 sets Variable Bit Rate quality 2 (approximately 190 kbps).

VBR quality 2 is widely considered 'near-transparent' — meaning even trained listeners on high-end equipment struggle to distinguish it from the original source. It's the quality setting used by most professional podcast and music distribution workflows. The resulting MP3 file is small enough to email, upload to Discord, or store on a phone, but high enough quality for any consumer playback scenario.