Extract Audio from Video
Pull the soundtrack out of any video file and download it as a WebM/Opus audio file — free, and entirely inside your browser.
How does this audio extractor work?
Your video is decoded by the browser and its audio output is routed through the Web Audio API into a recording destination. The tool then plays the clip through once at normal speed while the browser's own Opus encoder captures the audio into a WebM container. Because the whole graph is built in the tab, there is no upload, no queue, and no server — the download is ready the moment playback reaches the end of the file.
GuideHow to use it
Features & benefits
Why this runs entirely in your browser
Most online audio extractors ask you to upload the entire video, wait in a processing queue, and trust that the file is deleted afterwards. None of that is necessary. Modern browsers can already decode video and encode audio on their own, so this tool simply connects those two capabilities locally. Your video stays on your device, the extraction starts instantly regardless of file size, and it keeps working even without a network connection.
Common uses
Saving the audio of a lecture, podcast, or interview you recorded on video, lifting a voiceover or sound effect out of a clip for reuse, keeping the audio from a meeting recording so it can be transcribed, or turning a video you own into something you can listen to without the video track taking up space.
Frequently asked questions
A few things people usually want to know before extracting audio.
The download is a WebM file containing an Opus audio track — the only audio encoder every modern browser exposes to a web page without shipping a large extra library. Opus is an excellent codec that typically sounds better than MP3 at the same bitrate, and WebM audio plays in every current browser, in VLC, and in most editing software. MP3 is not offered because its encoder is not built into browsers; if you specifically need MP3 or WAV, convert the downloaded WebM afterwards with a desktop converter.
Extraction happens in real time because the audio is captured from live playback rather than being demuxed from the container. The browser decodes the video, routes its audio through the Web Audio graph, and encodes what passes through, so a ten-minute clip takes about ten minutes. The upside is that this approach needs no multi-megabyte WebAssembly download, works with every codec your browser can play, and starts producing output immediately instead of after a long upload.
Yes — leave the "Play audio out loud while extracting" option turned off and the capture runs silently. Internally the video's audio is routed into a recording node but not connected to your speakers, so the signal is still fully captured while nothing is audible. This is the right choice in an office or on a call. Turning the option on simply also connects that same signal to your speakers, which is useful when you want to confirm the correct audio track is being captured.
Any video your browser can decode will work, which covers MP4/H.264, WebM, and most MOV files. Codecs your browser cannot play — some HEVC or professional formats — will not load at all. If a video genuinely has no audio track, the extraction still completes but produces a silent file, so check that the clip has sound before starting. Videos with multiple audio tracks will capture only the default track, since that is the one the browser plays.
No. The video is never uploaded, stored, or shared with anyone. It is read locally from your device through an object URL, decoded by your browser, and encoded back out by the browser's own MediaRecorder — every step happens inside this tab, and no part of the file is transmitted over the network. The extracted audio exists only in memory until you download it, and closing or refreshing the page discards both the video and the result immediately.