Library workflow
Organizing Patreon and Kickstarter STL drops
By Petros, who builds Meshory · Updated Sep 18, 2026 · 8 min read
A Patreon subscription is a standing order for files. Subscribe to six creators and the drops land every month, named after their campaigns, and the pile grows whether or not you ever open it. This guide is about the intake rather than the backlog: a routine that takes about ten minutes per drop and leaves the library usable a year later. If what you have right now is a Downloads folder with thousands of loose files in it, start with [the recovery guide](/organize-stl-library/downloaded-files-cleanup) and come back once the floor is clear.
Most advice about organizing 3D print files assumes a one-time cleanup: you have a mess, here is how to dig out. A subscription is a different shape of problem. The drop arrives on the first of the month whether or not last month's is filed, so the question is not how much effort the cleanup takes once. It is how much effort it takes every month, twelve times a year, forever. Anything over about ten minutes gets skipped, and one skipped month becomes a backlog.
Why creator drops are awkward specifically
Three things make subscription bundles harder to file than a model you downloaded one at a time from Printables.
- They arrive as archives, not files. A drop is a ZIP, RAR or 7Z with the creator's own folder scheme inside it, and that scheme never matches yours. Two creators will file the same kind of thing three levels apart.
- The same model ships several times inside one drop. Presupported and unsupported, hollowed and solid, lite and full, sometimes split for a smaller bed. These are near-identical geometry under different filenames, which is exactly the case a filename search cannot see.
- Context lives next to the model and gets separated from it. Release notes, printing instructions and the creator's own renders are usually loose files in the same folder. Move the STLs out and the instructions stay behind.
A filing scheme that ignores those three is really just a plan to re-download things you already paid for.
The ten-minute intake routine
The routine earns its keep on one test: six months from now, can you answer the question do I already own a suitable dragon without opening a single archive?
- 1Download into one inbox folder with one subfolder per creator. Not per campaign, not per month: per creator. The creator is the thing you will remember.
- 2Leave the archive sealed. Do not extract it. The next section is entirely about why.
- 3Index the drop so its contents are visible and searchable without unpacking. This is the step that turns an opaque archive into something you can browse.
- 4Tag the drop once, while you still remember what it was: creator, and roughly what the campaign was about. Two tags beat a folder hierarchy you will have to maintain.
- 5Run a duplicate check before the next drop lands, not after five have piled up. A month of drops is a small enough batch to make decisions about.
Step four is the one people skip, and it is the one that pays. A tag applied in the ten seconds after a download is worth more than an hour of retrospective filing, because right then you still know what the thing is.
Leave the archives sealed
The reflex on receiving a ZIP is to extract it. For a subscription library that reflex is expensive. Extracting means you now hold the archive and its unpacked contents, so the drop costs twice the disk it needs to, and the unpacked copy arrives in the creator's folder scheme rather than yours. Delete the archive to save the space and you have lost the one artifact that proves what the drop originally contained.
The alternative is to index the archive where it sits. Meshory reads ZIP, RAR, 7Z and TAR/TGZ as though they were folders, rendering a thumbnail for each model inside and making it searchable, without unpacking anything. The creator's photos and any PDF instructions in the same archive open inline too, so the context stays attached to the models it describes. For a library that is mostly unopened campaign drops, this is the difference between doubling the disk they occupy and leaving them exactly as they downloaded.
An honest limit: reading inside an archive is not the same as printing from one. When you actually want to slice a model, it still has to come out of the archive first. What in-place indexing removes is the need to unpack the other 49 models you were not going to print this week.
File by creator, add meaning with tags
The instinct is to file by category: terrain here, busts there, mechs somewhere else. It falls apart the first time a drop contains all three, which for terrain and miniature creators is most drops. A folder can only put a model in one place, and a campaign bundle genuinely belongs in several.
Creator is the stable key. You subscribed to a person or a studio, their drops arrive under their name, and you will remember that name when you are hunting. Everything else that matters, the genre, the scale, the fact that you have already printed it, is a tag or a collection layered on top. The three-layer system covers that split in full; for a subscription library the short version is that folders track where a file came from and tags track what it is for.
The part that actually saves money
Subscribers re-buy models more often than they admit. A creator re-releases a remastered version, a bundle on a marketplace includes something from a campaign you already backed, or you back a second Kickstarter from the same studio that reuses sculpts. Filenames will not connect any of these, because every creator names things differently and re-exports change the bytes.
Matching on the geometry does connect them. Identical files group by content hash even inside archives, and the same mesh saved under a different name or converted to another format groups by geometry hash. Presupported and unsupported copies of one sculpt usually differ enough to count as separate models, which is correct, and a slicer project typically re-centers the model, so it surfaces as a likely duplicate for review rather than an exact match. The mechanics and the limits are in hash versus geometry detection.
| What you have | Filename match | Geometry match |
|---|---|---|
| Same file, re-downloaded under a new name | Misses it | Exact content match |
| Same sculpt, re-exported by the creator | Misses it | Identical geometry |
| Same model, opened and saved as a slicer project | Misses it | Likely duplicate, flagged for review |
| Presupported and unsupported versions | Sometimes catches it | Treated as different models, which is right |
If the backlog is already the problem
Everything above assumes you are starting the routine from today and letting the past sit. That is deliberate, because an intake habit is worth more than a cleared backlog and is much easier to start. Once the monthly drop stops adding to the pile, the pile itself is a finite job you can do on a wet Sunday. What to do with 1,000+ downloaded STL files is that job, in order.
For which software handles any of this, the organizer roundup covers the free options first and is honest that most libraries do not need a paid tool. Archives and geometry matching are the two places where the free tools genuinely stop, which is why they are the two things this page leans on.
Patreon library questions
What is the best way to organize Patreon STL files?
File by creator rather than by category, leave the archives packed and index them in place, and tag each drop once as it arrives. Category folders break the first time a drop contains terrain, busts and mechs together, which is most drops. Creator names are the thing you will actually remember six months later.
Should I unzip my Patreon downloads?
Not by default. Extracting doubles the disk each drop takes and hands you the creator's folder scheme instead of yours. Index the archives in place so their contents are searchable and rendered, then extract only the model you are about to print.
How do I stop re-buying models I already own?
Match on geometry rather than filenames. A creator's remaster, a marketplace bundle that overlaps a campaign you backed, and your own re-export all produce different bytes and different names for the same sculpt. Content and geometry hashing group them; a filename search never will.
How many models does a Patreon subscription actually produce?
It varies by creator, but a mid-sized miniature or terrain campaign often ships 20 to 60 models a month once presupported and unsupported variants are counted. Six subscriptions can therefore mean over a thousand models a year, which is why the intake routine matters more than the annual cleanup.
Do I need special software for this, or will folders do?
Folders and a naming convention handle the filing. What they cannot do is show you what is inside a sealed archive, or tell you that two differently named files are the same mesh. If your library is mostly unopened bundles, those two gaps are the whole problem; if it is mostly loose files you have already looked at, folders are genuinely enough.
Does indexing an archive modify it?
It should not, and in Meshory it does not: archives are read, never rewritten, and nothing is unpacked or moved on your behalf. That matters more than usual here, because a campaign archive is often the only record of what a drop originally contained.
Point it at the folder your drops land in
Meshory indexes ZIP, RAR, 7Z and TAR/TGZ where they sit, renders the models inside, and finds the copies you already own. Nothing is moved, unpacked or uploaded.