Background Audio Remover

Remove background noise and music from any video or audio file — a free background noise remover and audio cleaner that keeps the voice.
Free in your browser — no sign-up, no watermark. Cloud is optional.

Fast: wind, hum, fans, traffic and room noise go, every voice stays.

Process

Minutes from $4.99 · Pro $14.99/mo unlimited · Pricing

Drop a video or audio file

or

Choose file

MP4 · MOV · WEBM · MKV · MP3 · WAV · M4A · AAC · FLAC · OGG · OPUS

Loading noise model…
No file handy?
  • No upload — files stay on your device
  • Original deleted from the cloud after processing
  • Free, files up to 3 hours
  • Free 30-second preview · 3 free minutes when you sign in
  • No sign-up, no watermark

Working on your device · nothing is uploadedProcessing in the cloud · original deleted after processing

Cleaning your file…

0%Starting…
  1. Reading the sound from your file
  2. Removing background noise
  3. Saving the clean sound into your file

Keep this tab open until it finishes — closing it stops the work.Keep this page open, or leave your email below and close it. The address of this page brings you back to this job.

Done · cleaned on your device

Your file is clean

Download

Switch between Original and Clean while it plays — the picture keeps going, only the sound changes. Press S to switch.

Processed in the cloud. Your original has been deleted there; this result can be downloaded for 24 hours.

0:00 / 0:00
Removed
Voice warmthVoice sounds thin? Add back some low tones.
Download

Still hear music?

Email me the download link

Open it on another device or save it for later. The link works for 24 hours.

Only if you ask: the finished file (never your original) is uploaded to our storage and deleted after 24 hours. Your address is used once to send the link and is not stored. Delivered by Resend.The file is already stored on our side for 24 hours. Your address is kept with the job only until the link is sent, then deleted. Delivered by Resend.

How to remove background audio from a video

01

Drop your file

Any video from your phone, camera, screen recorder or video call. It starts on its own — there is no button to find.

02

Wait while it cleans

The noise is removed in your browser, with a progress bar and the time left. Nothing is uploaded.

03

Listen, then download

Switch between Original and Clean while it plays, then save the same file with clean sound.

What is a background audio remover?

A background audio remover — some people call it an audio background remover — takes the sounds you don’t want out of a recording and leaves the voice. This one does it with speech models: they have learned what voices sound like, keep them, and turn everything else down.

Unlike most online tools, it works on the video itself. You drop an MP4, MOV, WebM or MKV and get the same video back with clean sound; the picture is copied without re-encoding. Audio files (MP3, WAV, M4A, FLAC, OGG) work too. There is no account, no watermark, and you can hear Original and Clean side by side before you download.

A background noise remover first — music only when you need it

Above the drop zone you choose what goes. Most files only need the first option.

  • Remove noise (the default) — wind, hum, fans, traffic and room noise. Fast: it runs in your browser, usually quicker than the file plays, and nothing is uploaded.
  • Remove music — for videos with talking over music: the music goes, the voice stays. On the result, slide Keep other sounds to 0 for the voice alone, or up to 100 to keep footsteps, traffic and room sound too; you hear it at once. For speech, not for splitting a song into vocals and instrumental. More about removing music from video

“Remove music” takes longer, about 2 minutes or more per minute of sound. It runs locally on your device or, if you choose, in our cloud.

Hear the difference

Speech recorded next to a street: wind, a fan, 60 Hz hum, traffic and a car horn. Press play, then switch between Original and Clean.

0:00 / 0:15

Demo · public-domain LibriVox speech + noise we generated · cleaned by this tool, no other editing

Hear the music removed

A man reading over upbeat background music, with traffic behind him. Clean = “Remove music” with “Keep other sounds” at 0: just the speech is left. Press play, then switch between Original and Clean.

0:00 / 0:15

Demo · public-domain LibriVox speech + CC0 music (“Chillin’ Poupi” by Komiku) + traffic we generated · “Remove music”, “Keep other sounds” at 0, no other editing

What this video background sound remover takes out — the voice stays.

Wind noiseOutdoor vlogs, interviews, drone and action-camera clips. Hum & buzzElectrical 50/60 Hz hum, buzzing lights, ground-loop noise.
Fan & ACThe steady hiss of fans, air conditioning and computers.
Traffic & streetCars and city rumble behind vlogs, interviews and calls.
Room noise & hissBackground hiss from cheap mics, rooms and video calls.
Voice recordingsPodcasts, voice memos, lectures and voice-overs. Background music“Remove music”: the music goes, the voice stays — other sounds too, as much as you like.
Everything but the voice“Remove music” with “Keep other sounds” at 0: music, noise and every other sound go.

Not for: echo or reverb. Other people talking are kept in every mode.

Your file stays on your device

Most noise removers upload your video to a server. This one doesn’t. Your browser downloads the noise-removal model once (about 12 MB, then it is cached) when you open the page — or, in Chrome on Android with mobile data or Data Saver on, only when you add a file. “Remove music” uses a second, larger model (about 54 MB), which is downloaded only when you run it on your device. When you drop a file, your browser reads it, processes the sound and writes the new file — all on your own computer or phone.

For “Remove music” you can also choose Cloud (our processing server, reached through Cloudflare) instead, which is faster on phones. Then, and only then, your file is uploaded, processed in the cloud and deleted after processing (within 2 hours after a free preview); the result (and the voice and other sounds as two separate tracks, for the “Keep other sounds” slider) is kept for 24 hours so you can download it. Noise removal never uploads anything.

  • The models: DeepFilterNet3, an open-source speech enhancement model, for noise; Bandit v2, an open model that splits speech, music and other sounds, for “Remove music”. Both run as WebAssembly in your browser; if you choose “Cloud”, Bandit v2 runs there instead, with the same settings.
  • The picture: copied into the new file exactly as it was. It is not decoded or re-compressed, so there is no quality loss and it is fast.
  • The sound: processed, then saved at the same bitrate as your original sound, so the new file is about the same size.
  • Nothing to delete: there is no account. With “Local” your original is never uploaded; with “Cloud” it is deleted after processing (within 2 hours after a free preview). If you ask us to email you a download link, the finished file is stored for 24 hours and then deleted.

What you get back

  • The same kind of file. MP4 stays MP4, MOV stays MOV, WebM stays WebM, MP3 stays MP3, WAV stays WAV — M4A, AAC, OGG and FLAC too.
  • A clear voice, quiet pauses. Noise is removed during speech and in the gaps between words.
  • Music removed too, if you want. “Remove music” takes the music out; then the “Keep other sounds” slider sets how much of the rest stays, from the voice alone to the voice with traffic and footsteps. Singing counts as music and goes with it.
  • Voice warmth, if you want it (with “Remove noise”). Noise removal can make a voice a little thin. Voice warmth (Light, Medium, Strong) adds low tones back only while someone is speaking, so the noise doesn’t return in the pauses. Off is the default.
  • No watermark, and long files are fine. On your device, “Remove noise” takes files up to 3 hours long on a computer (30 minutes on a phone), “Remove music” up to 60 minutes (10 on a phone), and up to 4 GB (1 GB on a phone). Before a long run on your device, the page shows how long it will take there (and in the cloud, for “Remove music”), so you can choose. Cloud takes files up to 20 minutes (Pro: 60) and 100 MB.

Comparison

Looking for an Adobe Podcast Enhance alternative?

See how this tool and Adobe Enhance Speech compare on video files, accounts, length limits and privacy.

Questions

Does it keep the voice?

Yes — keeping the voice is the point. The models are built for speech: they keep every voice in the recording, including other people talking, and turn down what is not speech (with “Remove noise”, background music only partly). With “Remove music”, the “Keep other sounds” slider decides whether sounds such as footsteps and traffic stay too.

How do I remove background noise from video files?

Drop the video on this page (or tap Choose file). “Remove noise” is already selected, so it starts on its own. When it’s done, switch between Original and Clean while it plays, then download — you get the same video file with the noise removed and the picture untouched.

Is it free?

Yes. Everything runs free in your browser with no sign-up and no watermark; on your device, files can be up to 3 hours long on a computer (30 minutes on a phone). Cloud processing is optional: every file gets a free 30-second preview, new accounts get 3 free cloud minutes, and cloud minutes start at $4.99. Pricing

Will my video lose quality?

The picture doesn’t: it is copied into the new file as it is, without re-encoding. The sound comes back in the same format at about the same bitrate (WAV as 16-bit). The new sound is mono (the same on both channels), so stereo effects in the original are not kept.

Which file types does it accept?

Video: MP4, MOV, WebM and MKV. Audio: MP3, WAV, M4A, AAC, FLAC, OGG and Opus. You get back the same type of file: an MP4 stays an MP4, an MP3 stays an MP3.

Can it remove background music?

Yes. Pick “Remove music” above the drop zone. A second model (Bandit v2) splits the sound into speech, music and other sounds and leaves out the music. On the result, the “Keep other sounds” slider sets how much of the other sounds stays: 0 for the voice alone, 100 for the voice with footsteps and traffic. You hear each change at once, without processing again. Singing in the music counts as music and is removed. It is much slower than noise removal: about 2–3 minutes per minute of sound on an 8-core computer, longer on phones — or about 2 minutes per minute if you choose “Cloud” (a free 30-second preview, then cloud minutes). More about removing music.

What kind of noise can it remove?

With “Remove noise”: background noise behind a voice — wind, fans and air conditioning, electrical hum and buzz, hiss, traffic and room noise. Background music is removed only partly there; “Remove music” removes it. No mode removes echo, and other people talking are always kept.

Does it work on iPhone and Android?

Yes: Chrome on Android, and Safari 26 or newer on iPhone and iPad. A phone has less memory, so on a phone files can be up to 30 minutes long (“Remove music”: 10 minutes) and 1 GB; use a computer for longer ones. “Remove music” is much slower on a phone, so on phones it starts on Cloud (a free 30-second preview, then cloud minutes) — you can switch to “Local”, which is free.

Is my file uploaded anywhere?

Not with “Remove noise”, and not when “Remove music” runs on “Local”: then all processing runs inside your browser and your original is never sent to a server. If you choose “Cloud” (our processing server, reached through Cloudflare) for “Remove music”, your original is uploaded, processed and deleted right after processing (within 2 hours after a free preview, so the whole file can run without a second upload); the result is kept for 24 hours so you can download it. Cloud gives every file a free 30-second preview; the whole file uses cloud minutes (new accounts get 3 free ones). “Local” is free. And if you ask us to email you a download link, the finished file is stored for 24 hours.

How long does it take?

It depends on your device and the mode. “Remove noise”: in our tests a 26-second clip took about 10 seconds on an 8-core Android phone; computers are usually faster. “Remove music”: about 2–3 minutes per minute of sound on an 8-core computer, about 6 minutes per minute in browsers that can use only one CPU core, longer on phones; in the cloud about 2 minutes per minute, plus any wait in line. The progress bar shows the time left.

Why is the new sound mono?

Both models work on one channel. The result is put on both channels, so it plays normally everywhere. For speech this sounds the same; stereo effects in the original are not kept.

What does Voice warmth do?

Noise removal can make a voice sound a little thin. Voice warmth (with “Remove noise”) adds low tones back while someone is speaking and keeps the pauses quiet. Try Light first; Off is the default.

Updated 2026-11-03