On-device AI — professional-grade, fully private.

Tutorials

Split Audio into Parts, Every Way: Silence, Chapters, Fixed Length

Every way the free Audio Splitter cuts, on one 1:33 recording holding three takes: equal parts (three pieces of 0:31.2, each a byte copy of the original — 610 KB, nothing re-encoded, one ZIP), fixed length (30 s makes four, the last keeps 0:03.6), silence (the defaults find 24 gaps — every breath; a 2-second minimum gap finds exactly the three takes), and a pasted chapter list that names each piece, with a filename pattern. Choosing M4A re-encodes, and the tool says so. Nothing is uploaded.

Length: 3:01Published:

What you'll learn

Equal parts: 3 pieces, byte copy, one ZIP

Fixed length: the last piece keeps the rest

Silence: why the minimum gap matters

Chapter list: each piece takes its name

Filename patterns: {n} - {title}

When the format forces a re-encode

In this video

  1. 0:00Three takes, one file
  2. 0:29A byte copy
  3. 0:54Silence and the minimum gap
  4. 1:28A chapter list
  5. 1:47When it re-encodes
  6. 2:04Three jobs
  7. 2:21Free and Pro
  8. 2:34Nothing is uploaded

Every way the free Audio Splitter cuts, on one 1:33 recording holding three takes: equal parts (three pieces of 0:31.2, each a byte copy of the original — 610 KB, nothing re-encoded, one ZIP), fixed length (30 s makes four, the last keeps 0:03.6), silence (the defaults find 24 gaps — every breath; a 2-second minimum gap finds exactly the three takes), and a pasted chapter list that names each piece, with a filename pattern. Choosing M4A re-encodes, and the tool says so. Nothing is uploaded.

The Audio Splitter, every mode, on one recording that holds three takes — a voice note, a studio take and a meeting, 1:33.6 in all. Equal parts: three pieces of 0:31.2, and every piece is a byte copy — the original’s own MP3 frames written into new files, 610 KB each, nothing re-encoded — downloadable one by one or in a single ZIP. Fixed length: 30 seconds makes four pieces, the last keeping the 3.6 s that is left. Silence: at its defaults (-35 dB, 0.7 s) it finds 24 gaps — every pause for breath — and 13 pieces; raise "Gap at least" to 2 seconds and it finds only the breaks between the takes: three pieces, each cut in the middle of the break. Chapters: paste a time and a name per line and each piece takes its name; the filename pattern {n} - {title} makes "1 - Voice note.mp3". Choosing M4A re-encodes and says so; MP3 for an MP3 stays a byte copy. Free splits one file up to 50 MB and 30 minutes into 20 pieces; Pro takes 25 files and 500 pieces. Nothing is uploaded.

Full transcript

Every word spoken in the video, so you can read it, search it or skim it.

One recording · three takes

The Audio Splitter, every way it cuts: equal parts, a fixed length, the silences, and a chapter list — and why the pieces are never re-encoded. Free, and nothing is uploaded.

Equal parts · 3 pieces · 0:31.2 each

Byte copy · 610 KB each · one ZIP

Fixed length 30 s · 4 pieces · last 0:03.6

Silence · 24 gaps · 13 pieces

Gap at least 2 s · 3 pieces

Chapters · 3 pieces from your chapter list

{n} - {title} → 1 - Voice note.mp3

M4A → re-encoded · Same as source → byte copy

An album or a DJ set · Silence

A lecture for transcription · Fixed length

A podcast with chapters · Chapters

Free: one file · 50 MB · 30 minutes · 20 pieces

None of it left the browser. This is its own record of everything it fetched while all of that ran — not one request carried the audio.

Free, no account, nothing uploaded. Bring your own recording — and I will see you in the next one.

Subscribe for new tutorials

New walkthroughs every week. Free tools, nothing uploaded.