English · Português (Brasil) · Español · Français · Italiano · Deutsch · 日本語 · 简体中文

Bible: XML and JSON

141 Bible versions in 85 languages 🇧🇷 🇬🇧 🇪🇸 🇨🇳 🇸🇦 🇫🇷 🇩🇪

Clean, structured JSON and XML for Bible apps, verse search, and NLP/AI datasets.

📖 Just want to read? Try Open Bible, a free Bible reader built on this data. It works offline, installs as an app and needs no account. (source code)

Versions · File format · AI / LLM-ready · Wiki · Discussions

---

"And this gospel of the kingdom shall be preached in all the world for a witness unto all nations; and then shall the end come." — Matthew 24:14 (KJV)

🤖 AI / LLM-ready

Clean, per-verse structured JSON and XML — no HTML, no markup noise, no scraping artifacts. The same canonical book abbreviations are used across every language and version, so verses align out of the box for embeddings, RAG pipelines, fine-tuning, and cross-lingual alignment tasks. MIT-licensed code with permissively-sourced text, ready to feed straight into a dataset or a context window.

Do you want to create a Bible based app? What about an API for Bible verses search? Now you can do this, using sources in your very own language! The main objective of this project is to allow people to create Bible related apps using XML and JSON. Do you like this idea? Help us to go even further. Support this project! :)

How were the files built?

The source files were generated using a Python based crawler. Therefore, these files may contain minor issues related to encoding and syntax.

Which languages and versions are covered by the project?

Language	Version	Files	Notes
🇿🇦 Afrikaans	Afrikaans 1933/1953	af_afr53
🇬🇭 Akan	Akuapem Twi Contemporary Bible	ak_wakna
🇦🇱 Albanian	Bibla Shqip 1994	sq_albb
🇬🇷 Ancient Greek	Textus Receptus (Η Καινή Διαθήκη)	grc_grctr	New Testament only (27 books)
🇬🇷 Ancient Greek	Septuagint, Brenton (μετάφραση των εβδομήκοντα)	grc_grcbrent	Old Testament only (36 books), Greek Septuagint; Proverbs 30 is empty in this edition's numbering; Malachi has 3 chapters instead of 4 in this edition
🇸🇦 Arabic	Ketab El Hayat Majani	ar_kehm
🇦🇲 Armenian	Նոր Էջմիածին Աստվածաշունչ (New Etchmiadzin Bible)	hy_neb	Incomplete, 64 of 66 books (Esther and Daniel missing); Malachi has 3 chapters instead of 4 in this edition
🇮🇳 Assamese	Indian Revised Version (IRV) Assamese 2019	as_irvasm
🇪🇸 Basque	Navarro-Labourdin Basque Bible	eu_bhnt
🇧🇾 Belarusian	Bíblia (tradução de V. Semukha)	be_bbl
🇧🇩 Bengali	Indian Revised Version (IRV) Bengali	bn_irvben
🇻🇺 Bislama	Open Bislama Bible Revised	bi_obbr
🇧🇷 Brazilian Portuguese	Almeida Revisada Imprensa Bíblica	pt_aa
🇧🇷 Brazilian Portuguese	Almeida Corrigida e Revisada Fiel	pt_acf
🇧🇷 Brazilian Portuguese	Almeida Revista e Atualizada	pt_ara
🇧🇷 Brazilian Portuguese	Almeida Revista e Corrigida	pt_arc
🇧🇷 Brazilian Portuguese

Each entry's files are json/<code>.json and xml/<code>.xml (e.g. json/pt_nvi.json, xml/pt_nvi.xml). json/index.json lists this same catalog in machine-readable form.

How the files are structured?

Book names: since v3.0, name (JSON) and n (XML) hold the book name in the language of each version — for example Genesis in en_kjv and 創世記 in ja_jniv. Up to v2.3, almost every file carried the Portuguese name. abbrev is the stable identifier across languages, and json/book_names.json lists the names of every version. sa_san-dn and fr_apee keep English book names.

Some names are kept exactly as each edition publishes them and may look unusual: tr_ytc (Turkish) lost the dotless ı and some accents ("Yaratiliş" instead of "Yaratılış"); el_fpb and the grc_* editions have no accents and use σ instead of the final ς; nl_nld1939 calls Revelation "Apokalyps"; kk_qazbi (Kazakh) uses abbreviated names; lv_lg8 (Latvian) uses forms such as "Mateja ev."; and two names in om_onocv (Oromo, Genesis and Psalms) may be doubtful transliterations.

XML

The XML files are encoded using UTF-8 and built in the following structure:

<bible>
  <b id="abbrev" n="Book Name">
    <c n="1">
      <v n="1">Verse text</v>
    </c>
  </b>
</bible>

JSON

The JSON files are also encoded using UTF-8 and built in the following structure:

[
	{
	"abbrev": "abbrev",
	"name": "Book Name",
	"chapters":
		[
			["Verse 1", "Verse 2", "Verse 3", "..."],
			["Verse 1", "Verse 2", "Verse 3", "..."],
			["Verse 1", "Verse 2", "Verse 3", "..."]
		]
	}
]

You can get the number of the chapter and verse using the array index.

What about the license and the copyrights?

This project's code and data are distributed under the MIT License. All the Bible versions are property of their respective owners. All rights reserved to the owners.

How can I help the project?

You can review the code, enhance the structures or work on new versions. Every help is welcome! :)

Can I donate to the project?

Yes, you can! You can give a volunteer donation by PayPal.