CodeKitHubWhat Is This Tool?
Audio extraction is the process of separating the audio stream from a video container (like MP4, MOV, or WebM) without re-encoding the video — technically a "demux and transcode audio" operation rather than a conversion of the whole file. Because the video and audio streams are already stored separately inside the container format, the audio track can be pulled out directly.
This tool uses FFmpeg, the open-source multimedia framework that underlies most video and audio software on the market, compiled to WebAssembly via the ffmpeg.wasm project so it runs as JavaScript in your browser instead of needing a server or a native install. The first time you run it, your browser downloads the ~30 MB FFmpeg engine; after that, extraction happens entirely on your device using your own CPU.
Because there's no upload step, there's also no file-size limit imposed by a server, no wait for a queue, and no privacy question about where your video went — the same properties that make every other client-side tool on this site work the way it does.
Why Use It?
- You just finished a Zoom call recording and need only the audio for a meeting-notes transcription tool — drop the MP4 in, pick MP3, and skip opening a video editor for a two-minute task.
- You recorded a guitar riff on your phone as a video because that's the camera app you had open, and now you want just the WAV so you can drag it into your DAW without re-encoding loss.
- A client sent over a 200 MB product-demo video and you need to pull the voiceover for a podcast clip — no need to install Premiere or Audacity plugins for one extraction.
- You're archiving old family video files and want lightweight AAC audio backups of the commentary track without keeping every multi-gigabyte video around.
- You're on a shared or work computer where you can't install desktop software, but you still need to grab an audio track from a screen recording before a meeting.
- The source video contains a confidential client call or an unreleased demo track — processing happens locally, so the file never touches a server just to pull out the sound.
How to Use
- Click the drop zone (or drag and drop) to choose a video file from your device.
- Pick an output format: MP3 (small, widely compatible), WAV (uncompressed, best for further editing), or AAC/M4A (efficient, Apple-ecosystem friendly).
- Click "Extract audio" — the first run downloads the ~30 MB extraction engine, then processing happens locally.
- Preview the result with the built-in audio player, then click "Download" to save the file.
Example
Input
A 5-minute MP4 screen recording with narrationOutput
narration.mp3 (audio-only, same duration, no video track)The output audio's duration always matches the source video exactly, since extraction doesn't trim or re-time anything — it only removes the video stream and encodes the existing audio into the chosen format.
MP3 vs. WAV vs. AAC — which to pick
The right output format depends on what you're going to do with the audio next.
| Format | File size | Best for |
|---|---|---|
| MP3 | Smallest | Podcasts, voice notes, sharing, general playback |
| WAV | Largest (uncompressed) | Further audio editing, mixing, mastering — no quality loss to compound |
| AAC / M4A | Small, efficient | Apple devices/apps, good quality at a small size |
Why extraction is faster than a full conversion
Because the audio and video streams are already separate inside a container like MP4 or WebM, extraction only needs to demux the container and transcode the audio stream — the video frames are never decoded or re-encoded at all. That's why pulling audio out of a 500 MB video usually finishes in a fraction of the time a full video-to-video conversion would take on the same file.
- No video decoding: the tool skips reading and rendering every frame, since only the audio stream is touched.
- No video re-encoding: there's nothing to compress or resize, so CPU usage stays low compared to a video converter.
- Output size is tiny relative to the source: a 500 MB video's audio track is typically just a few MB to a few tens of MB depending on length and format.
Related tools
For other local media processing, try these tools.
→ Video Compressor (Target Size) · GIF Compressor · Image Compressor
Frequently Asked Questions
Is this the same as converting a video file?
Not quite — a video-to-audio "conversion" often implies re-encoding everything, but this tool specifically demuxes (separates) the existing audio stream from the video container and encodes just that stream into your chosen format, without touching or re-rendering any video frames.
Why does the first extraction take longer than the next one?
The tool needs to download the FFmpeg WebAssembly engine (~30 MB) the first time you use it in a browser session. After that initial load, the engine stays in memory and subsequent extractions start immediately.
Which format should I choose — MP3, WAV, or AAC?
MP3 is the smallest file and works everywhere, which is right for most cases like podcasts or clips. WAV is uncompressed and lossless, which matters if you're taking the audio into further editing where quality loss would compound. AAC/M4A is a good middle ground with strong quality-per-byte, especially common in the Apple ecosystem.
What video formats can I use as input?
Any format your browser can read metadata from and that FFmpeg supports, which covers essentially all common formats: MP4, MOV, WebM, MKV, AVI, and more. If a specific file fails, try a standard MP4 or WebM export instead.
Is my video uploaded to a server?
No. The video file never leaves your device — it's read directly into your browser's memory and processed entirely client-side using WebAssembly. Only the one-time FFmpeg engine download comes from a CDN; your actual video and the extracted audio never do.
Is there a file size or length limit on the video I can use?
There's no hard limit set by the tool itself — the practical ceiling is your device's memory and CPU, since everything is processed locally. Multi-gigabyte or hour-long videos will take noticeably longer and use more RAM than a short clip, but there's no server-side upload cap to worry about.
Does the extracted audio lose quality compared to the original video's audio track?
Choosing WAV preserves the audio losslessly, so there's no additional quality loss beyond what was already in the source. MP3 and AAC re-encode the audio into a compressed format, which introduces some quality loss — usually not audible at standard bitrates, but noticeable if you plan to edit or heavily process the track afterward.
Can I extract audio from a video that has multiple audio tracks (like different languages)?
This tool extracts the video's default/primary audio track. If a file has multiple embedded audio tracks (common in some MKV files), it will not let you choose between them — you'd need a dedicated video editor for track selection first.
Does the video need an internet connection to process after the engine loads?
No. Once the ~30 MB FFmpeg engine has downloaded for your session, extraction runs entirely offline in your browser — you can disconnect from the network and it will still work.