RunLocalTools

Extract an Audio Track From a Video — Pick Any Language, Free & Private

Pull one audio track or language out of an MP4, MKV, MOV or WebM video and save it as its own file. Lossless or MP3/FLAC/M4A. Runs in your browser — nothing is uploaded.

What it does

Movies, shows and screen recordings often carry several audio tracks in one file: English and Spanish dubs, a director’s commentary, a 5.1 surround mix beside a stereo one. This tool lists every audio track in your video — language, title, codec, channel layout, bitrate and length — and lets you save the ones you want as separate audio files.

You can pick tracks by hand, or type a language preference such as eng, en and let the tool choose the right track in each of many files at once. Save a track exactly as it is (lossless, in seconds) or convert it to MP3, M4A, OGG, Opus, FLAC or WAV.

How it works

Copy mode (lossless). An MP4, MOV, MKV or WebM file is a container: an index plus interleaved packets of video and audio. The tool reads the index, finds the packets that belong to the track you chose, and writes them into a new audio-only file without decoding or re-encoding anything. The sound is bit-for-bit what was in the video, and because only the audio bytes are read, an hour-long film is done in a few seconds.

The output container is matched to the codec so players accept it: AAC becomes .m4a (or raw .aac), MP3 stays .mp3, Dolby Digital becomes .ac3, Dolby Digital Plus becomes .eac3, FLAC becomes .flac, Opus is wrapped in an Ogg file as .opus. Anything else — DTS, TrueHD, Vorbis, PCM and more — goes into a Matroska audio file (.mka), which can hold any codec.

Convert mode. When you need a different format, the tool first copies the chosen track out losslessly (so only the audio, never the whole video, is handled), then converts it with ffmpeg compiled to WebAssembly, running in a background thread inside your browser. You can set quality, sample rate and channels (including a proper stereo downmix of 5.1/7.1), even out loudness to the EBU R128 broadcast standard (−16 LUFS), change volume, and add fade-in and fade-out.

Trim. Optionally enter a start and end time to export only part of a track. In copy mode the cut lands on the nearest audio packet (about 20–40 ms); in convert mode it is cut exactly.

How to use it

  1. Drop one or more videos onto the page (or click to choose them). The track table appears within moments, even for multi-gigabyte files.
  2. Tick the track(s) you want. Not sure which one is English? Press Preview to hear a 30-second sample from any point in the track.
  3. Optional: type your preferred languages (for example eng, en) and press Apply rules to all files to pick one track per video automatically — skipping commentary and audio-description tracks if you wish.
  4. Choose Copy (lossless) or Convert, and set a start/end time if you only need part of the audio.
  5. Check the file-name template — the default is {name} - {lang} - track {n} — then press Extract.
  6. Download each result, or all of them as one ZIP. Prefer the command line? Copy equivalent ffmpeg commands gives you the same extraction as text.

If you plan to tag the results as podcast episodes, add cover art and episode titles afterwards in the podcast ID3 tag editor. If a dialogue track comes out quieter than the rest of your library, the voice audio normalizer evens out the level, and when you have several extracted clips to join, the audio merger combines them with optional crossfades. To grab pictures from the same video, use the video frame extractor.

Supported formats and limitations

Video and audio files you can open: MP4, M4V, MOV, 3GP, M4A (including fragmented MP4 from streaming downloads), MKV, MKA and WebM.

Lossless copy works for: AAC (including HE-AAC), MP3, MP2, Dolby Digital (AC-3), Dolby Digital Plus (E-AC-3), FLAC, Opus, Vorbis, DTS, Dolby TrueHD, ALAC and PCM.

Convert outputs: MP3, M4A (AAC), OGG (Vorbis), Opus, FLAC and WAV.

Not supported:

  • AVI, MPEG transport streams (.ts/.m2ts), FLV, WMV and DVD/VOB files.
  • Encrypted (DRM-protected) tracks, such as downloads from streaming services. These are detected and shown as protected.
  • Converting very long surround tracks in one go: the converter has a memory ceiling, so a 5.1 track of an hour or more should be trimmed, downmixed to stereo, or saved with Copy mode instead. The tool tells you when a conversion is too large rather than failing.
  • Trimming an MP3 in copy mode can leave a faint blip at the very first frame, because MP3 frames can borrow data from the frame before. Use Convert mode for a perfectly clean cut.

Browser notes: the preview plays whatever your browser can decode. Many browsers (Chrome included) cannot play Dolby Digital directly, so for those tracks the preview offers to render a sample with the converter. Extracting is unaffected.

Frequently asked questions

How do I find the right language when a track is labelled “Undetermined”?

Many files carry no language tag, or a wrong one. Press Preview on each track: the sample player lets you jump to any point in the track, so you can skip past silent intros and listen for the language you want. The table also shows channel layout, bitrate and flags such as commentary, which usually separate the main mix from extras.

Is copying a track really lossless?

Yes. In copy mode nothing is decoded or re-encoded: the audio packets are moved out of the video unchanged. The saved file decodes to exactly the same sound as the track inside the video. Quality only changes if you choose Convert and a lossy format such as MP3.

When should I use Convert instead of Copy?

Use Copy when your player or editor accepts the track’s own format (M4A, AAC, MP3, FLAC and Opus are widely supported). Use Convert when you need a specific format — for instance MP3 for a car stereo, WAV for an editor, or a stereo MP3 from a DTS 5.1 track that many devices cannot play.

Can it handle a 4 GB movie file?

Yes for opening and copying: the tool reads only the index and the audio bytes in large sequential chunks and never loads the whole video into memory. A one-hour file with a 5.1 Dolby Digital track copies out in a few seconds on a typical computer. Converting is limited by browser memory — see the limitations above.

Will it work on a video I downloaded from a streaming service?

Only if the file is not copy-protected. Tracks encrypted with DRM are detected and cannot be extracted; this tool does not remove protection.

Does my video leave my computer?

No. See the privacy section below.

Privacy

Everything happens inside your browser tab. Your video, the extracted audio and your settings are never sent to a server, and the tool has no accounts, ads or tracking of your files.

What is local, and what touches the internet:

  • Reading the video, listing tracks, Copy mode, trimming and preview of browser-playable tracks: fully local, and they work offline once the page has loaded.
  • Convert mode (and previewing tracks your browser can’t play): the first time you use it, the page downloads the ffmpeg audio engine (about 25 MB) from the jsDelivr CDN. Only the engine is downloaded; none of your files or settings are uploaded. Your browser caches it, so later conversions need no download.
  • “Download all (.zip)”: loads a small zip-making library from the cdnjs CDN the first time. Again, nothing of yours is sent.

Saved in your browser: so a refresh doesn’t lose your work, small videos (up to 200 MB) and your finished results are kept in your browser’s own storage (IndexedDB) on this device only. Larger files are kept only if you tick the box on that file. Press Clear saved files & results at any time to delete everything the tool has stored.