Dataset Viewer
Auto-converted to Parquet Duplicate
The dataset viewer is not available for this split.
The number of columns (1105) exceeds the maximum supported number of columns (1000). This is a current limitation of the datasets viewer. You can reduce the number of columns if you want the viewer to work.
Error code:   TooManyColumnsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

eBible Parallel Corpus

A verse-aligned parallel corpus of 1253 Bible translations across 997 languages, prepared from publicly redistributable texts hosted by eBible.org.

Generated: 2026-05-04

Dataset Description

Each row corresponds to one verse reference from the standard 41,899-verse reference list (vref.txt). Every translation occupies one column, identified by its translationId. Rows are positionally aligned: row N in every translation column refers to the same verse.

  • Verse count: 41899
  • Translations: 1253
  • Languages: 997

Dataset Structure

Data Files

File Description
main.parquet Wide-format parallel corpus: one row per verse, one column per translation
metadata.parquet Per-translation metadata: language, licence, coverage, versification

Fields in main.parquet

Column Type Description
vref string Verse reference, e.g. GEN 1:1
<translationId> string Translated text for that verse; "" = untranslated; <range> = verse merged into preceding range

Fields in metadata.parquet

Column Description
translationId Unique identifier matching column names in main.parquet
languageCode ISO 639-3 language code
languageName Language name in the vernacular
languageNameInEnglish Language name in English
dialect Dialect, if applicable
title Full title of the translation
shortTitle Abbreviated title
description Description of the translation
textDirection ltr or rtl
script Writing script (e.g. Latin, Arabic)
inScript Whether the text uses the language's native script
OTbooks / OTchapters / OTverses Old Testament coverage
NTbooks / NTchapters / NTverses New Testament coverage
DCbooks / DCchapters / DCverses Deuterocanonical coverage
inferred_versification Versification number inferred for this translation
licence_Licence_Type Licence type (e.g. Creative Commons)
licence_Licence_Version Licence version (e.g. 4.0)
licence_CC_Licence_Link URL to the CC licence deed
licence_Copyright_Holder Copyright holder
licence_Copyright_Years Copyright year(s)
licence_Translation_by Translating organisation or individual
licence_Vernacular_Title Title as stated in the licence file
Copyright Copyright statement from eBible.org
UpdateDate Date last updated on eBible.org
sourceDate Date of the source text
publicationURL Publication page on eBible.org

Loading the Dataset

from datasets import load_dataset

ds = load_dataset("your-org/ebible-parallel", split="train")

# Access a specific translation
kjv = ds["eng-engKJV"]

# Filter to New Testament
nt_start = ds.filter(lambda row: row["vref"].startswith("MAT"))

Licence Breakdown

All translations in this dataset are redistributable as confirmed by eBible.org. Individual licence terms vary by translation. For per-translation licence details — including licence type, version, CC link, copyright holder, and translating organisation — load metadata.parquet.

Citation

If you use this dataset in your work, please cite David Baines, SIL, PABNLP:

@misc{ebible,
  title  = {eBible Parallel Corpus},
  author = {David Baines SIL Global PABNLP},
  url    = {https://pabnlp.org/},
  year   = 2026-05-04
}
Downloads last month
76