Speed up an audio file
without changing the voice

Yes, and there is nothing to sign up for. The work happens on your machine, so it costs us no server time and there is nothing to charge you for.

No. It is read from your disk into this page, changed here and written back out here. It never crosses a network, which you can check by pulling your cable out and using the tool anyway.

The recording is cut into windows of about fifty milliseconds. Each one is slid until its waveform lines up with the tail of the one already written, then cross-faded into it. Overlapping the windows more makes the file shorter without touching a single pitch period. Simply reading the samples faster, which is what most exports do, multiplies the pitch by the same factor and turns a voice into a chipmunk.

Anything your browser can decode: MP3, WAV, M4A, AAC, OGG and FLAC. Video files work in Chrome and Safari as well, and only their audio comes out.

Up to 250 MB, which is several hours of an MP3. A long file takes a few seconds of your processor because everything happens on your machine rather than on a queue somewhere.

Yes. Anything under 1x stretches the file out, with the pitch held in exactly the same way. 0.75x is the usual answer for transcribing an interview or following a lesson.

Choosing WAV re-encodes nothing: the samples come out exactly as they went in, only shorter. MP3 is compressed here at a normal listening bitrate, which is the same trade every MP3 makes.

On speech most people settle between 1.25x and 1.5x: the words stay easy to follow and a forty minute recording becomes twenty-six. 2x is comfortable once you are used to it, and 3x is really only good for scanning.

Yes, on iPhone and Android, in the same browser tab. A long file simply takes longer there, because the work is done by the device in your hand.