Find Quality Non-Fiction Books Like a Library Browsing the Stacks: The Book Prize Index Guide

⬅️ Back to Tutorials

Benjamin Breen, a historian at UC Santa Cruz and writer of the Res Obscura newsletter, spent his college years shelving books in Columbia’s Butler Library. He would flip to a random page of every book he shelved, sometimes getting absorbed, then reaching for the books on either side. It was a filtered serendipity: the Library of Congress system and the librarians had already curated the shelves, and the fact that a book had been checked out meant someone thought it was worth reading.

He built a digital version of that experience. The Book Prize Index is a free, searchable catalog of over 7,000 prize-winning non-fiction books. He had Claude Code scrape the data from Wikipedia and used semantic search so you can browse by feel rather than by keyword. Type “books for dads who like Pavement” or “David Attenborough but in book form” and the embedding model surfaces actual titles.

TLDR:

Browse 7,000+ non-fiction books that won or were shortlisted for major English-language prizes (Pulitzer, National Book Award, NBCC, etc.)
Use semantic search to find “books like” one you loved, or describe a vibe and see what surfaces
Explore by subject, publisher, imprint, or prize to find patterns in what gets recognized
Tools are free, no account needed. Breen pays hosting costs himself.

Reference: Quality non-fiction books are the antithesis of AI slop on Res Obscura, by Benjamin Breen (July 2026).

Why prize data works

Book recommendation algorithms are usually bad because they optimize for what sells, not what lasts. Amazon will recommend the same five self-published diet books to everyone. Prize lists are a filter that solves this: a jury of experts has already done the curation. These books have been read, debated, and judged by people who know the field.

The data covers 55 award programs and 121 prize categories, from the Pulitzers and National Book Awards to more specialized prizes like the Mark Lynton History Prize or the PROSE Awards. Each book gets a recognition score based on how many prizes it won and how many lists it appeared on, weighted by the prestige of each prize.

Semantic search changes how you browse

The most useful feature is the semantic search. It does not just match keywords in titles. You can describe what you want in natural language and the embedding model finds books whose descriptions and content match the feel of your query.

Try searching “long weird biographies of strange people” or “books about exploration that read like adventure novels.” The results are often surprising and genuinely good. This is the closest digital equivalent to pulling a book off a library shelf because the spine looked interesting, then pulling the one next to it because the topic turned out to be fascinating.

The keyword search mode is also there for when you need exact matches. Toggle between them in the search bar.

The publisher and imprint data

One of the most useful pages is the Publishers section, which ranks imprints and publishing houses by how many prize-winning books they have produced over time. You can see which small presses punch above their weight, how the big houses compare, and which imprints to watch if you care about quality non-fiction.

The subjects breakdown is also worth exploring. History leads with 1,554 books, followed by Biography (1,114), Arts and Criticism (681), and Science (614). Each one links to its full list.

Data experiments and visualizations

Breen built a few fun side projects with the data. The chromatic index arranges roughly 5,000 book covers by color. The trends page shows the rise and (potential) decline of non-fiction prizes over time, peaking around 2014.

What Breen argues about non-fiction

The Substack post that accompanies the tool makes a bigger argument: quality non-fiction books are the antidote to AI-generated sludge. They are carefully researched, peer-reviewed in practice if not in name, and they endure. A good non-fiction book from 1993 (like David Levering Lewis’s biography of W.E.B. Du Bois, which won four prizes) can be bought used for four dollars and read with as much profit as anything published this year.

Breen also traces the golden age of non-fiction to factors like jet travel enabling multi-continent research, the erosion of class and gender barriers in academia, and early digital cataloging standards. It is the kind of argument that makes you want to read more books, which is the whole point.

Why read the full post

The Substack post includes Breen’s personal reflections on library browsing, open stacks, the decline of research libraries, and what it felt like to build a tool using Claude Code to solve a problem he cares about. The Book Prize Index itself is immediately useful, but the framing (prize-winning non-fiction as a quality signal in an age of AI noise) is worth sitting with.