How to use Extract Audio from Video
- 1Drop in a video file
- 2Choose MP3 for a compact file or WAV for uncompressed audio
- 3Run the extraction — the video stream is discarded
- 4Wait while the audio is decoded and re-encoded
- 5Download the audio file
What Extract Audio from Video does
The audio inside a video is a separate stream, usually AAC in an MP4 or Opus in a WebM. Extracting it means decoding that stream and writing the samples into an audio file — the video track is discarded entirely and never touched.
Both output options re-encode rather than copy the original stream. That matters for how you should think about quality: WAV gives you exactly what the decoder produced, with nothing further discarded, but it cannot recover what the video's own compression already removed. MP3 adds a second generation of loss on top of that, which at 192 kbps is generally inaudible but is real.
When to use it
- A lecture or talk to listen to later
- Audio-only is a fraction of the size and plays in the background on a phone without keeping the screen on.
- Music or a soundtrack from a clip
- Pulling the audio out gives you something that plays in a music app rather than a video player.
- Feeding a transcription tool
- Most transcription services want audio rather than video, and uploading audio alone is far quicker.
- Reusing sound in an edit
- WAV output goes straight into an audio editor at full precision, which is where you want to start if the sound will be processed further.
Extract Audio from Video settings explained
- MP3
- Encoded at 192 kbps, a rate that is transparent for speech and close to it for music. The right choice for listening, sharing or transcription, and the file is roughly a tenth the size of the WAV.
- WAV
- Uncompressed 16-bit PCM. Nothing further is discarded beyond what the video's own audio compression already removed, which makes it the correct starting point for editing. Expect around 10 MB per minute of stereo.
- Video is discarded
- The picture stream is dropped before any encoding happens, so extraction is much faster than a video operation on the same file — only the audio has to be processed.
- Original track is not copied
- Both options decode the source audio and encode it again rather than lifting the compressed stream out intact. That keeps the output in a format anything can play, at the cost of one re-encode.
MP3 or WAV
| MP3 | WAV | |
|---|---|---|
| Compression | Lossy, 192 kbps | None |
| Size per minute | About 1.4 MB | About 10 MB |
| Further editing | Adds another generation | No further loss |
| Best for | Listening, sharing, transcription | Editing, mastering |
Limits and what to expect
- Both formats re-encode; neither copies the original compressed stream untouched.
- Quality is bounded by the video's own audio track — extraction cannot recover what was already discarded.
- Only the first audio track is extracted; videos with multiple language tracks give you one.
- Multi-channel surround is mixed down rather than preserved as discrete channels.
- The first run downloads a ~32 MB engine, which is then cached.
Troubleshooting
- The audio sounds worse than in the video
- For MP3, you are hearing a second generation of loss on top of the video's own compression. Choose WAV, which adds nothing further, or accept that 192 kbps is close to transparent for most material.
- The WAV file is enormous
- Uncompressed audio is about 10 MB per minute in stereo, and that is expected. Use MP3 unless you are going to edit the audio; for listening, the difference is not audible and the file is a tenth the size.
- "That video has no audio track"
- The file genuinely contains no sound — common with screen recordings made without audio capture enabled, and with video exported from some editors. There is nothing to extract.
- I got the wrong language track
- Only the first audio track is extracted. A file with several language tracks needs a tool that can select between them; this one always takes the first.
Your files and your privacy
Extraction runs through an ffmpeg build compiled to WebAssembly in your browser tab, on your own device. The video is never uploaded. That is worth checking against the alternative here, since sending a recorded meeting or lecture to a hosted converter means handing over the whole conversation.
What to use next
FAQ
Does extracting audio reduce quality?
WAV adds no further loss beyond what the video's audio compression already removed — it is exactly what the decoder produced. MP3 at 192 kbps adds a second generation, which is generally inaudible but is real. Neither can recover quality the video did not have.
Should I choose MP3 or WAV?
MP3 for listening, sharing or sending to a transcription service — 192 kbps is transparent for speech and the file is about a tenth the size. WAV if the audio is going into an editor, because starting from uncompressed samples avoids compounding losses.
Why is the WAV file so large?
It is uncompressed: every sample is stored in full, at roughly 10 MB per minute of stereo. That is the trade for adding no further loss, and it is why WAV is an editing format rather than a listening one.
Can I extract audio without re-encoding it?
Not with this tool. Both outputs decode and re-encode, which guarantees a file anything can play. Copying the compressed stream out intact would preserve it exactly but produce an AAC or Opus file that fewer things open.
What if the video has no sound?
You will be told the file contains no audio track. This is common with screen recordings made without audio capture and with video exported from some editors — there is genuinely nothing to extract.
Is my video uploaded?
No. Decoding and encoding both happen in your browser through WebAssembly, with nothing transmitted at any point.