Drop a recording and the waveform paints instantly. Set how quiet counts as silence and how long a gap has to be, watch the kept segments light up, preview the tightened result, then download a lossless WAV or a small MP3, all in your browser, never uploaded.
The Audio Silence Remover strips silent gaps and dead air from a recording. Drop an MP3, WAV, M4A or OGG, set a dBFS threshold, a minimum-gap length and how much breathing room to keep around speech, and the kept segments light up on a real waveform while a readout shows time reclaimed. Loudness is measured as windowed RMS like a real level meter, and every cut is crossfaded so the joins don't click. Preview the tighter cut, then download a lossless WAV or MP3: all processed privately in the browser.
The GigAI Audio Silence Remover decodes your file with the browser's native Web Audio API the moment you drop it, draws a crisp waveform of the whole take, then scans for silent stretches using two controls you actually understand: a threshold in dBFS that sets how quiet a passage must be to count as silence, and a minimum-gap length so a natural pause or breath isn't treated as dead air. As you drag those controls the detection re-runs on-device and the waveform lights up the loud segments it will keep, with a live readout of how many gaps were found and how much time you're about to reclaim. It's built for the real jobs people do: cutting the long silences out of a podcast take, tightening a rambling voice memo, removing the dead space between takes in a rehearsal recording, or shrinking a lecture where the speaker paused a lot. Two details do most of the work in making the result sound edited rather than chopped. A breathing-room control keeps a margin either side of every kept segment, because speech ramps up over a few tens of milliseconds and cutting exactly on the threshold shears the start off each word. And every join is crossfaded over a few milliseconds, which is what stops the cuts from clicking, butt-joining two unrelated waveforms steps the signal, and a step is an audible click. Detection itself measures loudness as RMS over short windows, the way a level meter does, so a single stray click in an otherwise silent passage no longer keeps the whole gap. When it looks right, preview the stitched result, then download a lossless 16-bit WAV (bit-for-bit accurate on every kept sample, ideal for further editing) or a compact MP3 for sharing: WAV is written natively and the MP3 encoder loads only if you pick MP3, so a simple silence trim never triggers a heavyweight download. All of it happens on your device: the file is decoded, scanned, cut and encoded in the tab, so confidential interviews, unreleased demos and personal voice notes never touch a server. There's no upload wait, no account and no watermark. This is honest gap removal. It deletes long silent spans and joins what's left. It is not a noise gate that ducks continuous background hiss, and it won't invent audio that was never recorded. Close the tab and every byte is gone.
Last updated
Drag an MP3, WAV, M4A, OGG or FLAC onto the workspace, or click to browse. It decodes locally in a moment and a waveform of the whole take appears with its length and channel layout.
Adjust the threshold (how quiet counts as silence, in dBFS) and the minimum gap length (how long a silence must be before it's removed). The kept segments highlight on the waveform as you tune.
Press preview to hear the stitched result. If a word got clipped, raise the breathing-room slider first, then try a lower threshold or a longer minimum gap. If too much dead air survives, do the opposite.
Pick lossless WAV or compact MP3 (and an MP3 bitrate if you like), then remove the silence, it's near-instant native Web Audio, not a WASM re-encode.
Check the before/after length and size, download the tighter clip, or reset and try different settings on the same file.
Set exactly how quiet a passage has to be before it counts as silence. Raise it to catch soft room tone, lower it to only cut true dead air. The detection re-runs instantly as you adjust.
A minimum-gap control means short breaths and beats between words survive, while only silences longer than your chosen length get removed, so speech stays natural, not choppy.
The loud segments the tool will keep are highlighted on a real waveform, with a live count of gaps found and seconds reclaimed, so you tune the cut by eye before you commit.
Hear the stitched, tightened result in the built-in player so you catch a clipped word or an over-aggressive cut before downloading. Not after.
Export a bit-perfect 16-bit WAV of the kept audio for further editing, or a small MP3 for sharing. WAV is native and instant. The MP3 encoder loads only when you choose MP3.
Decoding, silence detection, cutting and encoding all run in your browser. Interviews, voice notes and unreleased takes are never uploaded, stored or logged.
Instead of hunting for every gap by hand, one scan finds and removes them all, turning a rambling take into a tight edit in seconds, with every join crossfaded so the cuts don't click.
The highlighted waveform plus a live gap count and time-saved readout let you dial the threshold and gap length until the cut is exactly as tight as you want. No guessing.
The WAV path keeps your audio at full quality with no re-compression. Every kept sample is untouched apart from a few-millisecond crossfade at each join, which is there to stop the cuts clicking.
Because nothing leaves the tab, you can strip silence from a confidential interview or an unreleased demo without trusting it to anyone's server.
From quick one-off fixes to daily workflows, see how people put this tool to use.
Strip the long silences and mid-take pauses out of a raw recording before you edit, so you start your session with a tight timeline instead of scrubbing past empty space.
Remove the gaps where the speaker paused to write or think, turning a long recording into a shorter, denser listen you can get through faster.
Clean the dead space off the front, back and middle of a spoken take so the delivery feels tight, then export a lossless WAV for the mix.
Trim the gaps out of a confidential interview or a private recording entirely on your device, so the material is never uploaded to a stranger's server.
Accepts MP3, WAV, M4A, AAC and OGG, and produces WAV and MP3, all processed locally in your browser.
Trim audio in your browser with a real waveform: drop an MP3, WAV, M4A or OGG, drag a selection to keep exactly the part you want, preview it, then download a lossless WAV or a compact MP3. Sample-accurate and 100% private. Nothing is uploaded.
Cut a section out of the middle of an audio file in your browser. Drop an MP3, WAV, M4A or OGG, drag a selection over the real waveform to mark the part to delete, and the audio before and after is joined into one seamless clip. Download a lossless WAV or compact MP3. Sample-accurate and 100% private. Nothing is uploaded.
Split one audio file into multiple clips in your browser: by fixed-length chunks, by your own time ranges, or automatically on silence. Preview each part on its waveform, then download them individually or as a single ZIP. Sample-accurate, lossless WAV output, and 100% private. Nothing is uploaded.
Normalize audio volume in your browser: lift a too-quiet recording to a clean target level with peak or RMS normalisation, compare the before/after waveform, then download WAV or MP3. Nothing is uploaded.
Add a smooth fade-in and fade-out to audio in your browser with a live waveform. Drop an MP3, WAV, M4A or OGG, set the fade lengths, watch the volume ramp painted right on the wave, preview it, then download a lossless WAV or a compact MP3. Sample-accurate and 100% private. Nothing is uploaded.
Join audio clips into one track in your browser. Drop several MP3, WAV, M4A or OGG files, reorder them, hear the whole sequence play through, then download one seamless WAV or MP3. Gap-free, lossless on the WAV path, and 100% private. Nothing is uploaded.
Hit a snag? Here are quick fixes for the issues people run into most.
Either the recording has no silence long enough to qualify, or its quiet parts are still above your threshold. Raise the threshold (a less negative dBFS value) so softer passages count as silence, or lower the minimum-gap length so shorter pauses are removed.
The threshold is too high or the minimum gap too short, so speech tails and breaths were treated as silence. Lower the threshold toward a more negative dBFS value and raise the minimum gap. Re-check the highlighted waveform and preview before exporting.
The browser couldn't decode that codec natively, usually a niche or DRM-protected file. Try a standard MP3, WAV, M4A or OGG, or convert it first with the Audio Converter and then remove silence from the result here.
In-browser audio holds the whole decoded track in RAM while it scans and rebuilds. Close other heavy tabs and retry. If it still struggles, split the file into parts first, or run it on a machine with more memory.
Start with a moderate threshold and a 0.4–0.6s minimum gap, then tighten from there, it's easier to remove more than to recover a word you clipped.
Watch the gap count and time-saved readout: if it barely changes when you nudge the threshold, your recording's silence is either much quieter or much louder than you expected.
Keeping natural micro-pauses makes speech sound human, raise the minimum gap if the preview feels rushed or breathless.
For music or noisy rooms, lower the threshold so only true dead air is cut. Constant background will otherwise count as 'not silent' and stay anyway.
Export WAV while you're still editing (it's lossless), and only switch to MP3 for the final share-ready file.
Recent updates and improvements to the audio silence remover.
Initial release, drop-to-decode with a native Web Audio waveform, dBFS threshold and minimum-gap controls, kept-segment highlighting, stitched preview, and lossless WAV export with one-click download.
Added compact MP3 export with a selectable bitrate (encoder lazy-loaded only when MP3 is chosen), plus a live gap count and time-saved readout that updates as you tune the controls.
Improved silence-detection responsiveness on longer files, added a before/after length and size comparison on the result, and sharpened the honest guidance on gap removal versus noise gating.
Three quality fixes to how cuts are made. Joins are now crossfaded over a few milliseconds, so removed gaps no longer leave an audible click where two waveforms meet. A new breathing-room control keeps a margin around each kept segment so word onsets aren't sheared off. And detection now measures loudness as RMS over short windows rather than sample by sample, so a lone click in a silent passage no longer keeps the entire gap.
Your audio is decoded, scanned for silence, cut and encoded entirely in your browser using the native Web Audio API. It is never uploaded to any server. Nothing is stored, logged or transmitted, so you can safely de-silence confidential interviews, voice notes and unreleased recordings. WAV export is a lossless, native operation and the MP3 encoder is only loaded if you choose MP3. There is no cloud copy and no upload step. Close the tab and every byte is gone.
Free, private and instant. Audio Silence Remover runs right in your browser.
