No upload, 100% local, no account

Extract MP4 audio → MP3

Convert the audio track from your videos to MP3. Everything happens in your browser, no file is sent to a server.

How Extract audio from MP4 works

Extract the audio track from any MP4, MKV or WebM video and save it as an MP3, entirely in your browser. Drop your video file, and ffmpeg compiled to WebAssembly demuxes the container, decodes the audio stream and re-encodes it as MP3 using libmp3lame, all on your device. No video data is uploaded, and the download is ready as soon as processing finishes.

This is the right tool for archiving the audio from a talk you recorded yourself, pulling a music track from a video file, or converting a lecture clip to a portable MP3 for listening offline. The first run loads the ffmpeg WebAssembly engine, which takes a few seconds; subsequent conversions start immediately.

How to use Extract audio from MP4, step by step

  1. Drop an MP4, MKV, WebM or similar video file onto the upload area.
  2. Wait for the ffmpeg WebAssembly engine to load (first run only, a few seconds).
  3. Click 'Extract audio' and watch the progress indicator while ffmpeg demuxes and re-encodes the audio stream.
  4. Download the MP3 file directly to your device when extraction is complete.

Common use cases

  • Lecture audio: extract the audio from a recorded university lecture MP4 to listen while commuting.
  • Music archival: pull the audio track from a music video file you own and save it as a portable MP3.
  • Podcast repurposing: extract the audio from a recorded video interview to publish as a podcast episode.
  • Voice memo recovery: retrieve the audio from a screen recording or webcam clip where only the spoken content matters.

Frequently asked questions

Which video containers can I extract audio from?

MP4, MKV and WebM are the most commonly used and all work well. Other containers that ffmpeg supports (MOV, AVI, TS) also work in practice. The audio stream is decoded and re-encoded to MP3 locally regardless of the source container format.

Why does the first conversion take longer to start than subsequent ones?

The tool downloads and initializes the ffmpeg WebAssembly engine on first use. That engine is several megabytes in size, so the initial setup takes a few seconds depending on your connection. After the first conversion, the engine is already in memory and subsequent extractions start immediately.

Is the extracted MP3 the same quality as the audio in the video?

The audio is decoded from the video container and re-encoded to MP3 using libmp3lame. This means one generation of lossy encoding is applied. If the source video had AAC audio, the conversion is AAC to MP3, which introduces some quality change. For the highest possible quality, consider extracting to WAV using the audio-converter tool instead.

Does extracting audio from the video upload my video file to a server?

No. ffmpeg compiled to WebAssembly runs entirely inside your browser tab. Your video file is loaded into local memory, the audio stream is demuxed and re-encoded there, and the MP3 is offered as a local download. Your video never leaves your device. Open browser DevTools and check the Network tab: no request carries your file.

What if the video has multiple audio tracks?

By default ffmpeg selects the first audio track in the container, which is usually the main audio. If you need a specific track from a multi-track MKV, use the audio-converter tool with a custom ffmpeg command, or a dedicated desktop tool that lets you select streams.

Can I extract audio from a very large video file, such as a 2 GB recording?

In principle yes, but the entire file must be loaded into browser memory during processing, which requires your device to have enough RAM. Very large files may approach browser memory limits on devices with 4 GB of RAM or less. For very long recordings, splitting the video first with the video-converter tool reduces memory pressure.

Can I pull the audio out of a video on my phone?

Yes for a typical clip, a few minutes of video takes little memory since only the audio is re-encoded and the video stream is discarded. A very large multi-gigabyte recording still needs to be loaded whole into the browser tab first, which is where an older or lower-RAM phone struggles more than a computer.