KleioBase vs FamilySearch Full-Text Search: Searching vs Researching

FamilySearch Full-Text Search is the single most useful thing to happen to online genealogy in years, it costs nothing, and you should be using it. That is an odd way to open a page on a site that sells a competing-sounding product, so let us be precise about why it is not actually a competing product.

The short answer. FamilySearch Full-Text Search searches inside one enormous collection. KleioBase is a research companion for your whole family archive: it links the records you have, works out what is missing, and searches the open web for where those missing records are likely kept. Full-Text Search runs handwriting recognition across FamilySearch's own digitised images so you can search inside documents that were never indexed by name, but it cannot read a scan sitting on your hard drive. KleioBase holds no record images of its own and cannot search inside FamilySearch's collections for you. If the record you need is in a FamilySearch collection, search there. If you have forty records in a folder, no idea which ones describe the same person, and no clear idea where to look next, that is the other problem.

Full-text search solved the browse-only problem

For decades, a large share of digitised genealogy was browse-only. The images existed, nobody had indexed them, and finding your ancestor meant paging through a film reel by eye. Traditional indexing captured names, dates and places, which meant that everything else in a document, every witness, every neighbour, every property description, every marginal note, was invisible to search.

Full-Text Search changed the shape of that. It creates searchable transcripts of entire documents, so a witness who appears once in a deed becomes findable. As of early 2026 it covered nearly two billion record images, up from roughly 100 million at launch, spanning US land and probate records, Mexican notarial records, Brazilian civil registration, and hundreds of collections that had never been formally indexed.

That is a genuinely different capability from anything we offer, and it is free. The honest recommendation is to exhaust it before paying anyone, including us.

Its boundary is the collection

The limit is simple and it is not a criticism: Full-Text Search searches FamilySearch's images. It is a search engine over a specific corpus.

It does not take an upload. The photograph you took of your grandmother's naturalisation certificate, the scan a cousin emailed you, the church record a parish priest in Poland photocopied and posted to you, the ship manifest you bought from a regional archive: none of those are in the corpus, so none of them are searchable by this tool. For many families, and especially families whose records sit in archives that have not been digitised by anyone, that describes most of what they actually hold.

Language coverage is also still expanding rather than complete. It began with English, Spanish, and Portuguese, with Chinese, French, German, Dutch, and Italian announced for 2026. Coverage moves fast enough that you should check the current list on FamilySearch rather than believe a date-stamped claim on a comparison page, including this one.

Reading a document is an input, not the product

Worth stating plainly, because the shape of this comparison hides it: KleioBase is not a transcription tool. Transcription is one of two ways to get data in. The other is a GEDCOM import, and a researcher who imports a tree and never uploads a single scan still gets the whole product - duplicate detection across the tree, conflicting dates surfaced as conflicts rather than silently overwritten, missing generations flagged as gaps, a tree health score, and a companion that can be questioned about the whole archive in plain language.

That is the product. The reading step exists so that documents can join the same analysis, not because text is the deliverable.

Finding a record and understanding a record are different work

There is a second boundary that matters more than the first, and it is easy to miss because search feels like the whole job.

Suppose Full-Text Search works perfectly and hands you fourteen documents mentioning a Jan Kowalski. You now have fourteen documents and one question: how many men are these? Search does not answer that. It found matches on a string. Deciding that six of them are your Jan, four are his cousin, three are unrelated, and one is your Jan under a name the clerk spelled differently, is the analytical work that eats the actual hours in genealogy.

FamilySearch's shared tree is where many people then do that work, and it has a real structural feature: one tree, edited by everyone. That is a strength for collaboration and a well-known hazard for accuracy, which is why FamilySearch's own 2026 plans emphasise merge tooling and alerts before edits that look likely to degrade existing information.

Side by side

When FamilySearch is the better answer

You have not found the records yet. This is the common case and it is not close. Nearly two billion free images beats anything you can do with a folder of forty scans.

Your records are in a covered collection. US land and probate, Mexican notarial, Brazilian civil registration and the other covered families are searchable in ways that were impossible a few years ago. Use that.

You want collaboration. The shared tree connects you to living relatives working the same lines. A private archive by definition does not.

Your budget is zero. FamilySearch is free, permanently, and it is run by an organisation whose purpose is not to sell you anything.

What KleioBase does instead, and what it does not

We are a tool in this space, so treat this as disclosure.

KleioBase starts where your folder does. You upload the records you have gathered, from any source, in any of more than 30 languages, and it reads and translates each one, extracts the people, dates, places and relationships, and links them across every other record in your archive. The same ancestor in a baptism entry and a ship manifest becomes one profile rather than two scans with similar names. It flags likely duplicates so you can merge them deliberately, surfaces gaps where a generation or a spouse is missing, and lets you question the whole archive in plain language instead of rereading every document. When a gap turns up, the Research Companion can search the public web for the archives, record collections and finding aids most likely to hold the missing record, and link you to each one.

Crucially, it is private. Your archive is yours, not a contribution to a tree anyone can edit.

Now the honest limits.

We have no record collection of our own. The Research Companion can search the public web and suggest where a missing record is likely kept, with links to its sources, but it cannot sign in to FamilySearch or any subscription site and search inside their images. When the record you need is in a FamilySearch collection, Full-Text Search is still how you get to it.

We do not replace an archive's catalogue. We organise what you gathered. The finding aids of the institution holding the originals are still where you go to know what exists.

Everything is reviewed by you. Extraction is shown for confirmation before it enters the archive. That is deliberate, and it means a large batch carries real review time.

No DNA. We work from documents. Genetic matches are somebody else's product.

Where to start on your next research session

Run the FamilySearch search first, every time, and run it on more than the name. Search for a witness, a neighbour, a property description, a town. That is what full-text unlocked and it is the part most people still do not use.

Then look at your folder. If it holds fewer than a dozen documents, you do not have a linking problem yet and you should keep searching. If it holds forty, gathered over years, in more than one language, and you have ever opened two of them side by side to work out whether they describe the same man, that is the problem we built for. The knowledge base overview shows what that looks like in practice, and the Explorer tier is enough to try it on the records you already have.