Searching the Bible

BibleQL has three different ways to find verses, and picking the wrong one is the most common source of surprise. This page covers search; the other two have their own guides.

Choosing among the three

Query	Matching	Ordering	Use it for
search	Case-insensitive substring	Canonical	Quick lookup of a literal phrase
concordance	Stemmed word	Canonical, exhaustive	Word study, every occurrence
semanticSearch	Meaning, via embeddings	By similarity	"Verses about forgiveness"

How search matches

GraphQL

query {
  search(translation: "eng-web", query: "love", limit: 5) {
    bookName
    chapter
    verse
    text
  }
}

cURL

curl https://bibleql.org/graphql \
  -H "Authorization: Bearer $BIBLEQL_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{"query":"query { search(translation: \"eng-web\", query: \"love\", limit: 5) { bookName chapter verse text } }"}'

Ruby

require "bibleql"

client = BibleQL::Client.new(
  api_key: ENV.fetch("BIBLEQL_API_KEY"),
  api_url: "https://bibleql.org/graphql"
)

results = client.search("love", translation: "eng-web", limit: 5)

results.each { |v| puts "#{v.book_name} #{v.chapter}:#{v.verse}" }

Node.js

import { BibleQLClient } from "bibleql-js";

const client = new BibleQLClient({
  apiKey: process.env.BIBLEQL_API_KEY!,
  apiUrl: "https://bibleql.org/graphql",
});

const results = await client.search("love", {
  translation: "eng-web",
  limit: 5,
});

results.forEach((v) => console.log(`${v.bookName} ${v.chapter}:${v.verse}`));

Response

{
  "data": {
    "search": [
      {
        "bookName": "Genesis",
        "chapter": 22,
        "verse": 2,
        "text": "He said, “Now take your son, your only son, whom you love, even Isaac, and go into the land of Moriah. Offer him there as a burnt offering on one of the mountains which I will tell you of.”"
      },
      {
        "bookName": "Genesis",
        "chapter": 24,
        "verse": 67,
        "text": "Isaac brought her into his mother Sarah’s tent, and took Rebekah, and she became his wife. He loved her. Isaac was comforted after his mother’s death."
      },
      {
        "bookName": "Genesis",
        "chapter": 25,
        "verse": 28,
        "text": "Now Isaac loved Esau, because he ate his venison. Rebekah loved Jacob."
      },
      // ... 2 more
    ]
  }
}

search performs a case-insensitive substring match on the verse text. That has concrete consequences:

• love matches love, Love, beloved and lovingkindness — any verse whose text contains those letters.
• love does not match loved as a separate concept; it matches it only because loved contains the substring love. Conversely, searching loved will not find love.
• There is no stemming. Searching run will not find ran.
• There is no phrase or boolean syntax. The whole query string is one substring.

If you want linguistic matching — amor finding amores and amoroso — use concordance, which runs through the translation's stemming dictionary.

Results and limits

Results come back in canonical order (Genesis → Revelation, then chapter, then verse), not by relevance. There is no ranking.

The limit argument defaults to 25 and is capped at 100, server-side. Asking for 500 returns 100 without an error, so do not assume you received everything you requested.

search has no pagination. If you need every occurrence of a word rather than the first hundred, that is exactly what concordance is for — it is exhaustive and cursor-paginated.

Language behaviour

Substring matching is language-agnostic, so search works identically on every translation with no index required. That is its main advantage: no setup, no per-translation caveats.

It is also why it cannot do anything clever. Case is ignored, so amor does find Amor — but accents are not folded, so oracion will not find oración. If your users type without accents, normalize their input against the translation's own spelling before searching, or use concordance, which goes through a real text-search dictionary.

Practical guidance

• Looking for a phrase you can quote exactly → search
• Studying a word and its forms → concordance
• Describing an idea rather than wording → semanticSearch
• Need every match, not a sample → concordance