Datasets:

pykeio
/

librivox-tracks

license: cc-by-4.0
task_categories:
  - text-to-speech
  - automatic-speech-recognition
pretty_name: LibriVox Tracks
size_categories:
  - 100K<n<1M
language:
  - ace
  - afr
  - grc
  - ara
  - ban
  - bel
  - ben
  - ceb
  - bug
  - bul
  - yue
  - cat
  - zho
  - hrv
  - ces
  - dan
  - nld
  - eng
  - epo
  - fao
  - fin
  - fra
  - glg
  - deu
  - ell
  - heb
  - hin
  - hun
  - ind
  - gle
  - ita
  - jpn
  - jav
  - pam
  - kor
  - kur
  - lat
  - lav
  - lit
  - nds
  - ltz
  - mkd
  - zlm
  - mlt
  - mri
  - mar
  - enm
  - min
  - nor
  - nno
  - oci
  - ang
  - kaw
  - non
  - osn
  - tpw
  - ori
  - fas
  - pol
  - por
  - ron
  - rus
  - gla
  - srp
  - slk
  - slv
  - spa
  - sun
  - swe
  - tgl
  - tam
  - tur
  - ukr
  - urd
  - vol
  - wln
  - cym
  - fry
  - yid

A dataset of all audio files & corresponding sources uploaded to LibriVox before 1st November 2025.

Donate to Internet Archive, who hosts most of the audio data
Donate to Project Gutenberg, who hosts most of the source books

Be a good netizen and use proper caching & sensible rate limiting when downloading files.

November 2025 Update

book.id, section.id, and section.num are now integers instead of strings.
reader has been replaced with a readers array. No entries currently have more or less than 1 reader.
text has become source. The text field is now a direct link to the plaintext content of the book, if such content exists; null otherwise. 250K entries (~53%) have plaintext source available.