How to extract the audio from a video (MP3 or WAV)

Attach the video and ask Recut to "save the sound as an MP3" (or WAV). Recut converts it on your device with ffmpeg and puts the audio file next to the video, with no model cost.

Step by step

1. Attach the videoAdd an MP4, MOV, WebM or other video to a Recut project.
2. Say the formatMP3 is small and plays everywhere; WAV keeps the sound uncompressed for further editing.
3. Run itThe conversion runs on your device. Nothing is sent to an AI model.
4. Use it or save itSave the file, or keep going in the same project: transcribe it, clean it up or split it into stems.

What Recut does

Recut's convert step takes the sound track out of the video. It is also the first step Recut plans by itself whenever a later step needs audio rather than video, such as transcription, noise removal or separating vocals.

The same convert step changes formats in general: a video to MP4, WebM or GIF, a sound to MP3 or WAV, and an image to PNG, JPG or WebP.

Because nothing goes to a model, extracting the sound is free and works with no provider connected. Inside the browser it runs with ffmpeg compiled to WebAssembly, loaded the first time you need it.

Tips

• To replace a video's sound later, add the edited audio back with the video's own sound muted.
• To keep only part of the sound, ask for a trim with a fade in and out in the same plan.

What it doesn't do

• If the video has no sound track, there is nothing to extract and Recut will say so.
• Formats other than MP3 and WAV (such as AAC or FLAC files) aren't offered as outputs.

Related guides

• How to transcribe a video to textAttach a video and ask for a transcript: Recut takes out the sound and runs a speech-to-text model such as ElevenLabs Scribe or Whisper.
• How to remove background noise from a videoRecut takes the sound out of the video, isolates the voice from noise with an AI audio model, and puts the clean voice back on the picture.
• How to turn a video into a GIFAsk Recut to convert a clip to a GIF: 480 pixels wide at 12 frames a second by default, or the size and frame rate you say. Runs on your device.
• How to separate vocals from music in a songRecut runs a stem separation model, such as Demucs or SAM Audio, to split a song into vocals and instruments, or remove the vocals for karaoke.