Frequently asked, occasionally dignified questions.

Shelvd is professional collection management for physical books: antiquarian books, rare editions, signed copies, fine bindings, dealer stock, and the very ordinary books that somehow still deserve better than row 842 of a spreadsheet.

No. Modern books are welcome. Shelvd is simply built for the moment when title, author, and cover image stop being enough. Edition, impression, dust jacket, provenance, condition, valuation, inscriptions: all the details that make a copy your copy.

Goodreads tracks what you read. Shelvd catalogs what you own. One remembers that you finished a novel in 2018. The other remembers that your copy has a clipped jacket, a gift inscription, light foxing, and a valuation history best discussed calmly.

Dealers are very much invited. Shelvd supports professional descriptions, condition notes, image labels, export, catalog entries, valuation, provenance, and workflows for people whose stock has opinions, margins, and occasionally a deadline before a fair.

Yes. Collector is not a demo wearing a little paper hat. It covers up to 500 books with core cataloging, lookup, search, provenance, valuation, import/export, and trial AI scans.

No. One AI Scan credit covers one scan with 1 to 3 photos. Title page, cover, colophon: three photos, one credit, no tiny invoice hiding behind the dust jacket.

Start with the title page. It is the book’s passport, birth certificate, and sworn testimony. Add the cover for physical clues, and the colophon or copyright page for edition, printer, limitation, and the small print normal people wisely ignore.

Yes. That is rather the point. Books existed before barcodes, despite what some cataloging apps seem to believe. Shelvd can work from visible bibliographic evidence instead of demanding a number invented long after half your shelf was already in circulation.

AI Book Scanning is built for title pages in many languages, old imprints, Roman numerals, blackletter, Latin colophons, and printers who apparently charged punctuation by the handful. The rule remains simple: if the photo is readable, the AI has a fighting chance.

Up to 31 fields: title, subtitle, contributors, publisher, publication place, year, edition, impression, language, ISBN, colophon text, and more. It reads what is visible and leaves the invisible alone, which is a refreshing quality in software and people.

You review everything before saving. The AI is fast, useful, and occasionally overconfident, which makes it a junior cataloguer after espresso. You decide what stays.

AI Scan creates a new draft book from photos. AI Enrich revisits photos already attached to an existing book and suggests improvements. Scan is “please start this record.” Enrich is “please look again; we may have missed something expensive.”

AI Scan costs 1 credit per scan, up to 3 photos. AI Enrich costs 1 credit per selected photo, max 3. Existing books can have many images, and inspecting them à la carte is fairer than charging everyone for the full guided tour.

Collector gets 10 one-time trial scans. Collector Pro gets 10 scans per month. Dealer gets 50 scans per month. Enough to test the magic, feed a serious cataloging habit, or process a small mountain of newly acquired consequences.

No. Purchased credits stay until used. Monthly credits are spent first because they reset. Purchased credits are the sensible ones in the back room, waiting patiently with a ledger.

Yes. Packs come in 25, 50, 100, 500, and 1000 scans. For when “I’ll just catalog a few books tonight” becomes a weekend, a spreadsheet, and a mild change in personality.

The scan button stops politely. No overdrafts, no surprise charges, no dramatic billing ambush. You can buy more in Billing when you are ready to resume pointing a camera at bibliographic evidence.

They remain yours. Monthly credits stop when the paid plan ends, but purchased credits do not vanish in a puff of subscription logic. Shelvd is many things; petty is not the business model.

Use Library Lookup for canonical metadata from libraries. Use AI Scan for the copy in your hand: title page, colophon, signatures, condition clues, and the physical evidence no remote catalog record can politely inspect.

Yes. Condition is not a dropdown where “Good” goes to retire. Shelvd supports condition grades, notes, dust jacket condition, condition history, repairs, inspections, and the kind of specificity buyers and insurers appreciate.

Yes. Add owner chains, dates, evidence, inscriptions, bookplates, auction records, purchase prices, and the occasional gap where history went out for a walk and forgot to come back.

Yes. Shelvd tracks valuation history with appraisals, auction results, dealer quotes, insurance values, market research, and purchase prices. A book’s value is a story, not a single number wearing a tie.

Yes. Add photos and label them by book part: cover, spine, title page, colophon, damage detail, provenance evidence, and many more. A photo called IMG_4827 is a mystery. A labeled photo is evidence.

Yes. Import from spreadsheets with column mapping, then export in common formats including Excel, CSV, JSON, XML, RIS, BibTeX, and SQLite. We made it easy to leave because holding your data hostage would be rude and, frankly, unimaginative.

You do. Your catalog entries, provenance records, notes, valuations, tags, images, and collection metadata remain yours. Shelvd is the cabinet, not the deed.

No. Shelvd does not sell your data, does not sell your reading habits, and does not train AI models on your private collection data. The privacy policy says this plainly, without a footnote trapdoor.

They stay. Books above the free limit are not deleted, images remain accessible, and purchased AI credits remain in your balance. You may need to upgrade again before adding more, but the existing catalog does not get marched into the sea.

Yes. Account deletion exists because GDPR says it should and because sometimes people really do want to close the cabinet. Export first if you want a copy; deletion is permanent, as permanent things irritatingly are.

Shelvd is built in Belgium by Bruno van Branden under Simplinity. No venture capital theater, no faceless support labyrinth, just a collector with too many books and an unreasonably detailed database.