Semantic book search API
Big Book API's /search-books endpoint takes a plain-language query and parses it with its own ontology, so synonyms, hypernyms and homonyms are resolved without any query syntax. A search for "books about dogs" also finds books about border collies. Filters for genre, rating, publish year and author then narrow the results.
What the ontology resolves
- Synonyms
- Different words for the same thing match the same books.
- Hypernyms
- A broader term covers the narrower ones under it: "dogs" includes border collies and other breeds.
- Homonyms
- A word with more than one meaning is resolved to one meaning.
Write the query the way a reader would say it. There is nothing to switch on: every query to /search-books is parsed this way.
How it works
The documentation example searches for books about wizards:
curl -H "x-api-key: $BIG_BOOK_API_KEY" \
"https://api.bigbookapi.com/search-books?query=books+about+wizards" The response, abridged to two results:
{
"available": 6049,
"number": 10,
"offset": 0,
"books": [
[
{
"id": 17200576,
"title": "Journey to Hogwarts",
"image": "https://covers.openlibrary.org/b/id/2520429-M.jpg",
"authors": [ { "id": 14136020, "name": "J. K. Rowling" } ],
"rating": { "average": 0.8939999938011169 }
}
],
[
{
"id": 15954006,
"title": "Inside Hogwarts",
"subtitle": "Magical Paintbook",
"image": "https://covers.openlibrary.org/b/id/2520430-M.jpg",
"authors": [ { "id": 14136020, "name": "J. K. Rowling" } ],
"rating": { "average": 0.8939999938011169 }
}
],
…
]
} books is a list of lists; with group-results on, editions of the same book share an inner list. Ratings run from 0 to 1. Pass a result's ID to /{id} for the full record or to /{id}/similar for recommendations.
A query with filters
The filters are separate parameters, so the query stays in plain words. This asks for nonfiction books about dogs published after 2010 with a rating of at least 0.7, highest rated first:
curl -G "https://api.bigbookapi.com/search-books" \
-H "x-api-key: $BIG_BOOK_API_KEY" \
--data-urlencode "query=books about dogs" \
-d "genres=nonfiction" \
-d "min-rating=0.7" \
-d "earliest-publish-year=2010" \
-d "sort=rating" \
-d "sort-direction=DESC" \
-d "number=20" Search parameters
| Parameter | Example | Description |
|---|---|---|
query | books about wizards | The search query in plain language. |
genres | nonfiction | Comma-separated list of the 92 genres; books from any of them match. |
authors | J.K. Rowling | Comma-separated author IDs or names. IDs come from /search-authors and are faster. |
min-rating, max-rating | 0.8 | Rating range between 0 and 1. |
earliest-publish-year, latest-publish-year | 2022 | Books published after or before this year. |
isbn, oclc | 9781781257654 | Return only the book with this ISBN-13 or OCLC number (ISBN lookup). |
sort, sort-direction | rating, DESC | Sort by publish-date or rating, ascending or descending. |
group-results | false | Group similar editions of the same book. |
number, offset | 10, 0 | Return 1 to 100 books, skipping 0 to 1,000. |
Semantic search compared with keyword book APIs
The Google Books API takes one q string with field keywords such as intitle:, inauthor: and subject:, and returns at most 40 results per request. Open Library's Search API ranks by keyword relevance with no semantic or vector mode. Neither documents a way to expand a query to related concepts, so to be sure a search covers "dogs" and every breed, you write those terms into the query yourself.
| Big Book API | Google Books API | Open Library Search | |
|---|---|---|---|
| Query | plain language plus filter parameters | q with field keywords | keywords |
| Related concepts matched | yes, through the ontology | not documented (keyword syntax) | no, keyword relevance |
| Sort options | rating, publish date | relevance, newest | new, old, random, key |
| API key | yes | yes | no |
| Free use and quota | 50 requests/day, non-commercial | no published price; per-project daily quota, commonly reported as 1,000/day | 1 request/s, 3 with an identifying User-Agent |
Facts about other APIs come from their documentation as of November 2026; Google publishes no default quota, so that figure is the commonly reported one. Google's branding guidelines also forbid reordering its results and require a "powered by Google" graphic next to them, which matters if you plan to merge or re-rank results from several sources.
Limits and pricing
Each search counts as one request, whatever the number of results. The free plan gives 50 requests a day at 1 request per second, for non-commercial projects with a backlink. Paid plans cost $6, $29 and $99 a month for 500, 5,000 and 50,000 requests a day, then $0.04, $0.02 or $0.005 per extra request, at 2, 10 and 20 requests per second (pricing, quotas). The terms do not allow storing results; caching for up to an hour needs written permission.
When a keyword API is a better fit
- You need books in other languages or a larger catalogue. Big Book API covers over 4 million English books. Google Books has a
langRestrictfilter, and ISBNdb publishes a much larger catalogue figure. - You need to store the data. Open Library's catalogue data can be kept and reused, and it publishes monthly dumps.
- You want Google Books previews. The Google Books API can filter for books with partial or full previews.
Our book API comparison and Google Books API alternatives guide cover these trade-offs in more detail.
Frequently asked questions
What is semantic book search?
Search that works on the meaning of a query. Big Book API parses each query with its own ontology, so "books about dogs" also returns books about border collies.
Do I need query syntax such as intitle: or subject:?
No. Write the query in plain words and use parameters for the rest: genres, authors, min-rating, earliest-publish-year, sort and others.
How is this different from the Google Books API?
Google Books searches with keyword syntax in one q string and returns at most 40 results per request. Big Book API resolves synonyms and broader terms, returns up to 100 results per request and filters by rating, but covers only English books.
Can I combine a natural-language query with filters?
Yes. Every filter works together with query, for example genre, a minimum rating, a publish year range and a sort order in one request.
Does the semantic search work in other languages?
No. The catalogue has over 4 million books, all in English.
Try natural-language book search on the free plan: 50 requests a day.
Get a free API key Read the docs