Prusa Launches Their Own 3D Model Repository

If you own a 3D printer, you’ve heard of Thingiverse. The MakerBot-operated site has been the de facto model repository for 3D printable models since the dawn of desktop 3D printing, but over the years it’s fallen into a state of disrepair. Dated and plagued with performance issues, many in the community have been wondering how long MakerBot is still going to pay to keep the lights on. Alternatives have popped up occasionally, but so far none of them have been able to amass a large enough userbase to offer any sort of real competition.

Sorting models by print time and material required.

But that might soon change. [Josef Průša] has announced a revamped community for owners of his 3D printers which includes a brand-new model repository. While clearly geared towards owners of Prusa FDM printers (support for the new SLA printer is coming at a later date), the repository is not exclusive to them. The immense popularity of Prusa’s products, plus the fact that the repository launched with a selection of models created by well known designers, might be enough to finally give Thingiverse a run for its money. Even if it just convinces MakerBot to make some improvements to their own service, it would be a win for the community.

The pessimists out there will say a Prusa-run model database is ultimately not far off from one where MakerBot is pulling the strings; and indeed, a model repository that wasn’t tied to a particular 3D printer manufacturer would be ideal. But given the passion for open development demonstrated by [Josef] and his eponymous company, we’re willing to bet that the site is never going to keep owners of other printers from joining in on the fun.

That being said, knowing that the users of your repository have the same printer (or a variant, at least) as those providing the designs does have its benefits. It allows for some neat tricks like being able to sort designs by their estimated print time, and even offers the ability to upload and download pre-sliced GCode files in place of traditional STLs. In fact, [Josef] boasts that this is the world’s only repository for ready-to-print GCode that you can just drop onto an SD card and print.

Regular Hackaday readers will know that we’ve been rather critical of the decisions made by MakerBot over the last few years, but to be fair we aren’t exactly alone in that respect. The community desperately needs a repository for printable models that’s in somebody else’s hands, and frankly we’re thrilled with the idea it could be [Josef Průša] leading the charge. His printers might not be perfect, and they certainly aren’t cheap, but they definitely don’t fail to impress. Here’s hoping this latest venture will be the same.

58 thoughts on “Prusa Launches Their Own 3D Model Repository

    1. No joke. Sometimes its just easier to remodel a part then modify a god forsaken stl. Stl are not source code. Literally impossible to get accurate measurements off those terrible things.

        1. You mean the OpenSCAD program that’s a free download and is available on Windows, Mac and Linux? Yes, the CAD programs which run on only some operating systems, keep their file formats binary and secret and the ones that cost lots of money are WAY better than the free OpenSCAD.

          1. .STEP files are just about the most widely supported non-polygonal 3D file out there.

            It’s in the name (Standard for the Exchange of Product data)

            OpenSCAD is a pain to use, not intuitive, and not nearly as powerful as many other professional CAD packages.
            Just because something is open source, doesn’t automatically make it better than everything else.

          2. I agree that a cross product file format is great but my comment didn’t exclude that did it? I really don’t think this is the place to start another proprietary vs open source discussion. As a software developer and someone who can type with all 10 fingers with ease, typing out OpenSCAD designs is relatively easy. The few high school kids I taught it to had no problem getting basic designs and concepts using it. But it was a multi quarter class and they needed something done in a week.

            I would hope the Prusa model repository supports a variety of file formats and if any preferences are made, it’s toward formats which are easily shared between different products.

          3. .step is an open standard. Openscad models can not be sent to a machine shop for milling/turning, step files can. Even freecad will read in step files. You can even download step files for most mechanical parts on mcmaster carr.

            Any good cad package should import step parts, because it allows sharing mechanical designs between different packages easily. I do not have to learn some new workflow or cad tool with step files. I just download it into a cad package I know and get to work. Not waste my time downloading openscad, learning how it works, wasting times on forums to try to get it to work.

            I can run circles around Openscad users with modern cad packages when it comes to productivity.

          4. OpenSCAD is not optimized for complex models. Draw some 10cm thick plate and put 100×100 holes of 1mm thickness, and then try to rotate in browser.

            It doesn’t matter whether or not a part of the model is visible in the browser, OpenSCAD is trying to calculate the complete model without skipping invisible parts as real CAD’s doing.

    2. Have you checked out Wikifactory? (

      It combines version control, with visualisers (including STEP [0]), community tools (forum, issues, posts…), and we are looking into interfacing with production workflows soon.

      I’d be delighted if you gave it a go and let us know what you think!


  1. You can only upload gcode, which is specific to printers. Prusa printers. Unvetted and unsafe code that isn’t tuned to your printer with no way to change anything meaningful about it. If they were going for an alternative to Thingiverse they sorely missed the mark.

    And let’s not even discuss the new map feature either.

    1. You can upload STL files, and though it doesn’t say it on the page, you can also upload original source files, from sketchup, fusion, etc..

      I can see the appeal of gcode, assuming that all machines of a particular model act the same. Plus you could upload gcode to do experimental stuff not normally available in the slicers. Think uploading music, or anything else weird. Though personally I’d still prefer to slice the model myself.

      1. Even if all machines of a particular model act the same (in my experience, they do not, but close enough to make it work in general, but not perfect). Then still, uploading gcode is a bad idea, as material most certainly doesn’t act the same.
        Also, “intent” will not be the same between users. Some might want to print faster with less quality, some might want to maximize visual quality.

        Not to forget that there are about 50 popular 3D printer models out there, so uploading gcode isn’t sustainable if you want a decent user base.

        1. How many people will actually check the g-code for anomalies?

          How many printers run sanity checks on g-code for any potential problems?

          Those are the two largest issues with G-code, it creates a possible failure mode that would have been mostly eliminated through the natural process of running a 3d model through a slicer by the end user.

  2. I gave up on posting on thingiverse and youmagine years ago and warn 3D printing noobs at the makerspace to view any thingiverse post with a skeptical eye if there’s no photo of a printed part. I host my own files on my google drive now and put links in my own blog.

    It seems like there’s a lot of people who “design” things and post them to thingiverse because they want to have a large number of items on thingiverse, nevermind if they aren’t printable or don’t do what they are supposed to do. Maybe Prusa will filter out that type of design/designer.

    1. What would be nice is if there were a standard .json format that we could include in our own repositories that sites like Thingiverse, Youmagine and all others could scrape in to their own repositories so that no-one ultimately controls the availability of the models.

      1. It’s easy when you have a lot of experience, but not so easy for noobs- the main target of sites like Thingiverse- who don’t know what they are looking at or what a good design should to look like. Inexperienced users DL crap designs and find they can’t print them, and struggle unnecessarily to make a good print. Like my rustic grandmother used to say, “you can’t make a silk purse from a sow’s ear”. There are a LOT of sow’s ears on Thingiverse.

        What is needed is a curated site that filters out the crap. At the very least, requiring photographs of an actual print, reviewed by a human being, would eliminate 80% of the junk “designs”.

        Thingiverse is antisocial media (sort of like facebook) where any idiot with access to CAD can post any stupid thing and call it a “design”, and many inexperienced users easily fall victim to their trash.

    2. Worst thing is that theese suckers upload already existing models with cosmetic modifications which can rouin functionality. Also they mention briefly or don’t mention at all the origin, which is hard to find in the search result because of high number of clones.

  3. Github actually supports stls really well. We need some enterprising frontend engineer to help people upload stls, licences, and instructions there and glob them together in a place to find them all. Having storage and downloads on github, as well as their great stl and image viewers would no doubt save a lot of money on the ultimate frontend.

    1. Totally agree! I do a bunch of design in OpenSCAD, so github is a natural choice for .scad revision control. If it featured a frontend similar to Thingiverse for 3D designs, it could really make a dent in Thingiverse’s popularity.

      1. There was a lot of talk about starting up Thingiverse alternatives 8 years ago on #reprap. I remember a few people spun them up in various incarnations.

        The problem is the mass of files/users that Thingiverse has. That’s a huge hurdle, and not one that can be overcome with mere software. You’d have to offer something more/better to draw people away.

        What Thingiverse lacks is any sort of rating/feedback system. When you just want to print out a Millenium Falcon, you find 100 different models. Picking the best one is Herculean. Establishing which are printable at all is hard work. The “I made it” feature gets halfway there, but if you really nailed the curation aspect and kept it open, you would have a Thingiverse-killer, IMO.

        As @jeffeb3 suggests, you wouldn’t even need to host it yourself if you could leverage GitHub. You could just link to folks’ projects with a ratings/aggregation overlay.

        Go go go! I’ll buy you a beer.

        1. Thingiverse’s search function is also, in my experience, mostly garbage. The popularity of Thingiverse is as much a curse as a blessing – it’s HARD to find stuff and frequently what I’m looking for requires “odd” search terms that are nonintuitive.

          Partly because Thingiverse’s categories are WAY too broad. Prusa’s system seems to be a bit better organized into subcategories – although only a bit.

          “Hobby & Makers” -> “Electronics” is WAY too broad.

          About the only thing worse than Thingiverse for searching/organization is OSH Park’s “sharing” system.

          1. Yeah, the search function is abysmal. I’ve often been unable to hit on the proper set of search parameters to pull up designs I know are on the site. Often I’ll end up searching through Google instead, which ultimately points me to the right page on Thingiverse.

            Though I’ve also had to do the same thing to find articles I know have been posted here on Hackaday, so perhaps not much room to talk there.

      1. I have seen people use it enough times to know they find it easier to tweek a model via the customizer option than downloading OpenSCAD and opening up customizer locally. But you have your opinion.

  4. Sad so little (other then a few comments) mention ypumagine. Sonce its spin-off from ultimaker it is no longer bound to any printer and a small te of devs are working on it again.

  5. Implementing bitcoin as an exchange value system and collecting fees on a tiny scale (relative to the most stable know monetary system) may help keep the site up and bitcoin economy. Some sites have mining parking for browsers, which run at 20% gpu/cpu rate on an end client. And throw in some advertising and it will okay. May be move the site to MS Azure cloud and have it on stand by at Amazon AWS if the price is currently low.

    1. No, just no. Any site that tries to run bitcoin mining scripts on my computer gets blackballed immediately, Its so easy to tell when you run no-script.

      If a company needs to trick their users into mining bit coin for them, or they can only survive through advertisements then they really need to think about their business case and maybe pursue a different avenue. Yes i believe that most of the websites on the internet should die.

    1. If your thought was to just DL other people’s stuff to print, then probably, yes, you should rethink your desire to have a 3D printer. 3D printers are great things to have if you’re willing to learn to use them well, which includes making your own designs to print. That can take a lot of time and effort. If you’re looking at a low-budget printer, as so many noobs do, you may find you’ll spend as much time trying to get it to print as you will printing. You’ll also waste a lot of time trying to print unprintable files from Thingiverse.

    2. There’s still plenty of benefit of having a 3D printer. My suggestion to people who ask is: Find a friend who also is also about 1/2 into it, and share the cost of a printer. It brings the cost into the realm of a tinkering/learning tool and it gives you someone to share in the fun. Two people can solve problems 4x better than 1. You do not need to design things yourself to get a basic worth out of it, but there still will be a lot to learn and if you do end up designing things for yourself, you’ll have a low barrier of entry.

      TV has it’s problems, but just look at the most recent uploads and you’ll see it’s far from dead. Not to mention the hundreds of thousands of parts posted over the years.

  6. it’s literally on their main domain address …. it’ll never see remotely near the amount of use that even myminifactory gets. nevermind that it will confuse the average schmo (“but durh, i don’t own a prusa”) – but it also conveys that it’s a blatant marketing gimmic/tool and more about the brand than the community. More half-baked ideas poorly executed by Prusa. Always one step away from something cool, but never fully actualized due to silly ass and often identity driven mistakes.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.