How to Convert M4B to MP3 (and Keep the Chapters)

In your browser (no install)

1. 1

   Open the M4B to MP3 tool

   It converts in your browser, so the book is never uploaded.

2. 2

   Drop the M4B

   Its chapters, cover and details are read straight away.

3. 3

   Choose one file or one per chapter

   One MP3 keeps the chapter marks inside it; one per chapter makes numbered files any player can step through.

4. 4

   Pick a quality and convert

   Match or exceed the source’s bitrate; 64 kbps is fine for a 64 kbps audiobook. Download the MP3 or a ZIP of chapters.

With ffmpeg (Windows, Mac, Linux)

One MP3, keeping the chapters and cover

ffmpeg writes an ID3v2 tag with the cover as an attached picture.¹ When the M4B has chapters, it also writes one ID3 chapter frame per chapter plus a table of contents (tested with ffmpeg 6.1). Drop -map 0:v and the two -metadata:s:v options if the M4B has no cover.

ffmpeg -i book.m4b -map 0:a -map 0:v -c:a libmp3lame -b:a 64k -c:v copy \
  -id3v2_version 3 -metadata:s:v title="Album cover" -metadata:s:v comment="Cover (front)" book.mp3

One MP3 per chapter

This script reads the chapter times and titles with ffprobe and cuts each chapter to its own numbered MP3, titled and track-numbered. It handles commas and quotes in chapter titles, and -map_chapters -1 stops every file inheriting the full chapter list.² -q:a 4 is LAME’s default quality, which ffmpeg’s guide calls close to perceptual transparency; use -b:a 64k for small speech files.³

#!/usr/bin/env bash
# Usage: ./split.sh book.m4b  -> split/NN - Title.mp3, one per chapter
in="$1"; mkdir -p split
count=$(ffprobe -v error -show_entries chapter=id -of csv=p=0 "$in" | wc -l)
for ((i=0; i<count; i++)); do
  start=$(ffprobe -v error -show_entries chapter=start_time -of csv=p=0 "$in" | sed -n "$((i+1))p")
  end=$(ffprobe -v error -show_entries chapter=end_time -of csv=p=0 "$in" | sed -n "$((i+1))p")
  title=$(ffprobe -v error -show_entries chapter_tags=title -of default=nw=1:nk=1 "$in" | sed -n "$((i+1))p")
  n=$(printf '%02d' $((i+1)))
  ffmpeg -nostdin -hide_banner -loglevel error -y -i "$in" -ss "$start" -to "$end" \
    -map 0:a -map_chapters -1 -c:a libmp3lame -q:a 4 \
    -metadata title="$title" -metadata track="$((i+1))/$count" \
    "split/$n - ${title//[\/\\:*?\"<>|]/}.mp3"
done

Just one stretch

ffmpeg -i book.m4b -ss 00:10:00 -to 00:25:30 -map 0:a -c:a libmp3lame -b:a 64k ch02.mp3

Other ways

• VLC: Media, Convert/Save, then an MP3 audio profile. It converts the whole file to one MP3, without chapters.⁸
• Audacity: imports M4A and M4B once its optional FFmpeg library is installed, then exports MP3; chapters are not carried over.⁷
• Apple Music or iTunes: can convert tracks between formats, but Apple notes that converting between compressed formats can reduce quality.⁵

Why renaming doesn’t work

An M4B is AAC audio inside an MPEG-4 container. Changing the extension to .mp3 leaves the same bytes behind: a renamed file still reads as MPEG-4 AAC (we checked with ffprobe). Some players look inside and play it anyway; MP3-only devices reject it. Renaming to .m4a, on the other hand, is harmless, because M4A and M4B are the same format.