How to use SimilDoc
A guide for anyone, with no jargon. If you get stuck somewhere, write to us: anything unclear is the manual’s fault, not yours.
Before you start
SimilDoc finds documents that say almost the same thing and sorts them into folders. It is for when you have several versions of one file and no longer know which is the good one.
You need no account, no email and no internet connection. Nothing you do leaves your computer: there is no server on the other side.
It reads PDF, Word (.docx), plain text (.txt) and Markdown (.md). Anything it cannot read is listed by name, not quietly ignored.
The first time you open it
Since SimilDoc is not signed with a paid certificate, your system will warn you once that it does not recognize the author. That is normal for independent software and does not mean something is wrong.
- On Windows: if you see “Windows protected your PC”, click “More info” and then “Run anyway”.
- On Mac: unzip the file, drag SimilDoc wherever you want to keep it, and open it. If macOS blocks it, go to System Settings, Privacy & Security, and click “Open Anyway”.
It only happens the first time. If you want to be sure the file you downloaded is ours, its fingerprint is published next to each download so you can compare.
Step 1: pick a folder
Drag the folder onto the window, or click in the middle to browse for it. SimilDoc only looks at the folder you point it at.
If you want it to go into the folders inside too, open “Opciones avanzadas” and tick the subfolders option. Handy when a whole year is split across twelve monthly folders.
A tip for your first run: start with a small folder, a few dozen files. You get to see how it works before handing it your entire archive.
Step 2: find the similar ones
Press the search button. SimilDoc opens each document, reads its text and compares it with all the others. It does not look at names or dates: it looks at what they say inside.
A folder with a few dozen documents takes seconds. The first search of the day takes a little longer while the program finishes getting ready. If the folder holds thousands of files, go make a coffee.
Nothing is modified during this step. SimilDoc is only reading.
Step 3: review the groups
This is where you decide. Each group is a family of documents SimilDoc considers the same document in different versions, and each file shows how similar it is and when it was saved.
- The percentage
- How close that document is to its nearest neighbour in the group. 100% means the text is identical, even if one is a Word file and the other a PDF.
- The CLON tag
- That document says exactly the same as another one. SimilDoc treats the oldest as the original and marks the rest as copies, so you know which one to keep.
- The “Sin grupo” card
- Documents that match nothing else, and those that could not be read. They are not hidden: you see them too.
- The similarity slider
- If it grouped too much or too little, move it. To the right it demands nearly identical documents and makes fewer groups; to the left it is more permissive and makes more. The groups rebuild instantly, without reading anything again.
Up to here not a single file has moved. The bar at the bottom reminds you.
Step 4: sort them
Once the groups convince you, choose how it should work and press the organize button. There are two modes:
- Copy (recommended for your first run)
- Your original files stay exactly where they are. SimilDoc works on copies, so you can see the result without risking anything.
- Move
- It really tidies up: the files change place and your folder is left clean. Note that any shortcuts you had to those files will stop working.
What you will find afterwards
- A new folder called “SimilDoc - Organizado”, inside the one you chose.
- Inside it, one folder per family, named after the files themselves.
- A “Únicos” folder for what has no match, and a “No legibles” one for what could not be read. Nothing is left loose.
- Each file carries its saved date in the name, and exact copies say “(clon)”. Your originals are never renamed.
- A report in JSON with everything that was done. That is what makes undoing possible later.
If you change your mind: undo
Pick that same folder and the undo button appears. Every file goes back exactly where it was.
Undo follows the same rules: it overwrites nothing. If you put another file at the original path in the meantime, that one stays and SimilDoc tells you which documents it could not return, and why.
If you sorted in Copy mode there is nothing to undo: your originals never moved. You can delete the copies yourself whenever you like; SimilDoc does not delete files.
What happens to what it cannot read
Two cases end up in “No legibles”, always with the reason written down:
- Scanned documents
- A PDF that is a photo of paper has no text to read. SimilDoc does not recognize text inside images yet; it is on the list of what is coming.
- Suspicious or damaged files
- A document can be built to eat all your memory when opened. SimilDoc works out how much it would take before opening it and refuses it if that crosses the limit. It also skips files over 100 MB.
Questions that come up while using it
Can I close it halfway?
Yes. If you close it while it is analyzing, nothing happens: no file had been touched. If you close it while it is sorting, it waits for the operation in flight to finish, so nothing is left half done.
Does it work offline?
Yes, and that is the best way to check what we promise: disconnect and use it. It works the same, because it never needed the network.
What if I have thousands of documents?
It works, but it takes longer and uses more memory. If your folder is huge, go in batches: one year first, then the next.
Can I sort a folder I already sorted?
Yes. SimilDoc recognizes what is already organized and does not duplicate it. Each run leaves its own report, so you can always go back.
What about documents in the cloud (OneDrive, Dropbox, iCloud)?
It works the same, but reorganizing a synced folder makes your service upload those changes, like any other file you move. For your first run, try Copy on a small folder.
Does it group documents in other languages?
Yes. It compares the text as it is, so it works with any language written in letters. Accents do not get in the way: it reads “declaración” and “declaracion” as the same word.
If something goes wrong
Tell us what happened. Three things are enough: which system you use, what you did step by step, and what you expected to happen.
Do not send us your documents. We do not need them to fix a problem and we will never ask for them.
Ready to try it?
Free for personal use. No account, no email and, if you prefer, no install at all.