I Am Worried About Prusa Research

| Comments

I am a happy owner of a Prusa MK3S 3D printer. It has been in service here in my home office for nearly three years, and I haven’t had any major problems with it. I had a stretch last year where it was running for about five hours a day, three or four days a week for five months. Prusa Research really does make a workhorse of a 3D printer.

Look at that old thing!

Before that, I had been running an old MakerFarm Prusa i3 that I bought used in 2014, but it was starting to feel old, and it was starting to fail. Buying a fully assembled Prusa MK3S in 2020 was an easy choice. I could write 2,000 words explaining exactly why the Prusa MK3S was by far the best option for me at the time.

Would I buy a Prusa printer today?

I think about this a lot when talking about 3D printing on our Discord server. I wasn’t sure of the answer to this question until today.

I bought a Sovol SV06 last week. I can’t imagine that printer being my only 3D printer. It is cheap. It is loud. I don’t expect it to be terribly reliable. It was just so inexpensive that I just couldn’t say no.

They are the same printer!

I have been considering adding a Bambu P1P to my office for a while. The price is good. It is a much faster, much more modern printer than my Prusa MK3S. The Bambu P1P ships fully assembled at a price very close to a Prusa MK3S+ or Prusa MK4 kit. I don’t have much use for the AMS on the Bambu X1C, so the Bambu P1P seems like a no-brainer.

I always assumed that a Bambu P1P would have a home next to my proven workhorse of a Prusa. Even if my Prusa MK3S disappeared in some sort of calamity, I assumed that I would replace it with a similar machine from Prusa Research. I am not so sure I would do that anymore.

They are treating the Prusa XL like a slow, lumbering dinosaur

The Prusa XL looked extremely exciting when it was announced two years ago. Prusa Research was making a Core XY printer. It looks like a giant Voron! How awesome is that?! Vorons were printing rather nice benchies in 15 or less. It was exciting that we were going to get a fast, reliable Core XY machine from Prusa.

Except that isn’t what they are delivering. I just installed the Prusa XL profiles for PrusaSlicer. They are faster than the stock Prusa MK3 profiles, but they are far from fast.

I just installed a 0.6mm CHT nozzle on my Prusa MK3S. The Prusa XL ships with a 0.6mm nozzle. I have not finished tweaking my own profiles for speed yet, but I think I am at a reasonable point for now. I don’t have input shaping like a Voron or Bambu. My own 0.22 mm layer height DRAFT settings estimate 48 minutes to print a Benchy. The Prusa XL 0.25 mm layer height SPEED profile estimates 43 minutes.

Prusa MK3S Prints

I also sliced a 1x1x3 Gridfinity bin. With my 0.45 mm DRAFT profile, the estimate is 17 minutes. With the Prusa XL’s 0.40 mm DRAFT profile, the estimated print time is 24 minutes.

This is disappointing to me. The Prusa XL is a $1,999 printer. You can find videos of clean Benchy prints on the Bambu P1P in about 18 minutes. That is a $699 printer. You can slice the Benchy in Bambu’s ludicrous mode to get the print time down to around 12 minutes, but it won’t look as nice.

Vorons have been running with an input shaper for years. Marlin has an input shaper now. The Bambu printers have always shipped with an input shaper. This is how they get up to 20,000 mm/s² of acceleration, and that acceleration is how they print fast.

I limit my Prusa MK3S to 2,500 mm/s² when printing infill. Prusa Research limits the Prusa XL to 3,000 mm/s² when printing infill. We have this limit because the Prusa MK3S and the Prusa XL both currently lack an input shaper.

They missed their release date target for the Prusa XL by a year. The competition has been using input shaping during that entire time. How are they not shipping the Prusa XL and the Prusa MK4 with an input shaper?! Did they only just realize this week that people are craving faster printers?

The Prusa MK4 is shipping without an input shaper

This is the first weird. Josef Prusa sent out an email blast bragging about the Prusa MK4. The very first bullet point is bragging about their 20-minute Benchy. You need to use their input shaper to print a Benchy in 20 minutes.

Jo Prusa's Email

Guess what? They are shipping the Prusa MK4 without the input shaper. If you receive a Prusa MK4 today, you will not be able to print a Benchy in 20 minutes. They aren’t shipping that capability.

This is the very first bullet point. This is one of the things Prusa Research brags about in their intro video for the printer. The printer can’t yet do what has been advertised.

The Prusa MK4 is not yet open-source

This is also one of the bullet points in Jo Prusa’s email blast. He is bragging that the Prusa MK4 is open source. He said on his blog that they aren’t yet releasing the Prusa MK4’s control board schematics.

You should read his blog post about Prusa Research and open source. It left a bad taste in my mouth.

Open source is like a business partner that you don’t have to negotiate with. He built his company on top of the RepRap project, the Marlin project, and Slic3r. These all allowed him to get a jump start without also hiring a team to create firmware and a slicer from scratch in order to get his product out the door. He got a lot more from these open-source components in the early days than he does today.

Bender on my Sovol SV06

Open source is a two-way street, and he seems to be unhappy about that now that others are benefiting more than he is. Other companies can do the same things with his project as he was allowed to do with Marlin and Slic3r. Josef Prusa is in the position of power today, and he isn’t excited about sharing.

He seems to specifically be unhappy that there are exact copies of the Prusa MK3 kit on Aliexpress at half the price. I don’t know why he thinks these vendors need any form of source code to produce these copies, and I don’t know why he thinks a license change will make any difference.

You only need to buy one original Prusa MK3S kit to duplicate it. Most of the hardware is just off-the-shelf components like stepper motors, stepper drivers, and an LCD screen. The most complicated component is probably the controller, but that isn’t exactly rocket science to duplicate, and there are other open-source controllers that could be used in its place.

Trying to cancel their existing-customer discount

This one is a bummer, and it doesn’t make me feel good about the state of things. There was a composite of a bunch of screenshots posted on Reddit. I haven’t attempted to trace back to all the sources to verify that everything is completely accurate, but Josef Prusa left a comment in the thread, and he didn’t call any of it out as being fictitious. I am not a hard-hitting investigatory journalist, so that is good enough verification for me.

Here’s what the screenshot is explaining: The older Prusa i3 printers from the MK2 through to at least the MK3S listed a 5% or 10% discount on any future printer orders. That means my fully assembled Prusa MK3S entitles me to 10% off my next purchase of a Prusa i3.

Jo Prusa's response to the filament debacle

They don’t offer that discount any longer. Josef Prusa said on Reddit that he doesn’t want to offer this discount any longer, so they just took it away. Since being called out on it, he wants to “run it in a limited version in the future.”

I am sorry, but no. Absolutely not. As far as I know, when I bought my Prusa MK3S, part of the deal was that Prusa Research would be giving me a 10% discount on future Prusa i3 orders. This deal has already been made. You can’t just change the terms after money has exchanged hands. That is completely dishonest. I can think of some worse words to say about someone who tries to secretly change the terms of a previous deal.

Prusa Research isn’t some mom-and-pop shop that made a mistake that they can’t afford to buy their way out of. Forbes estimated that Prusa Research was worth €236,000,000 in 2016 based on €33,000,000 in annual revenue. Prusa’s website says they sold four times as many printers in 2021 as they did in 2018. Their blog says they sold over twice as many printers in 2021 as they did in 2018.

There is a gap in my data. I am sure it is available somewhere, but I think I have found enough information to prove my point that Prusa Research is a big business. They can afford to hire people to help prevent these kinds of pricing mistakes. They can also afford to pay for these mistakes after they happen.

What is the absolute worst case? Is it that Josef Prusa is somehow being scammed, and every single Prusa MK4 they ship out qualifies for the 10% discount? If that is the case, he has to raise the price of the Prusa MK4 by 11%. That is it. The fully assembled Prusa MK4 would have to cost $1,220 instead of $1,100.

We don’t know how many orders qualify for the discount, but it can’t possibly be every order. The worst case isn’t so bad, and the reality is likely much better.

Even so, I understand that Josef Prusa is in a tight spot. He has to compete with faster, larger printers that cost a lot less.

The Prusa MK4 comes with a spool of filament, except that it doesn’t, but now it does?

This is absolutely bonkers. What do you think it costs Prusa Research to manufacture a spool of filament? Maybe $10? Do they think there is no such thing as bad press?

I don’t think this apology would have rubbed me the wrong way a month ago.

He is apologizing for the confusion. Except no one is confused. The product page said the Prusa MK4 came with a 1 kg spool of PLA. People placed orders. Prusa Research decided they weren’t going to include 1 kg of filament with those orders. That isn’t confusion. That is a bait-and-switch.

Josef shouldn’t be apologizing for our confusion. He should be apologizing for changing the terms of the sale after the sale has already been made. This is turning into a pattern, and I don’t like it.

He says he is doing this to thank you for your preorder. No one should be expected to be thankful for receiving what they paid for. That is gross.

I like the technology in the Prusa MK4 and the Prusa XL!

I absolutely love the load-cell sensor for bed leveling on the Prusa XL and Prusa MK4. It isn’t a game changer like the PINDA probe or BL Touch were, but it is a really solid incremental improvement in quality of life. This is the upgrade I want more than any other, and I am excited to see what else they can do with this sensor in the future.

Bambu has been shipping equivalent technology for 11 months already. Instead of one load-cell sensor in the extruder, Bambu has three load sensors on the bed. I don’t know which is better, but both allow the printer to find the point where the nozzle makes physical contact with the bed to give us a perfect first layer.

It is exciting that Prusa is finally on board with input shaping and speeding up print jobs. They have been dragging their feet on that for years. I have always gotten the impression that Prusa Research doesn’t think people are interested in printing fast.

Bambu has been shipping a ridiculously fast 3D printer with input shaping for the past 11 months. Everyone I know is excited about how fast the Bambu printers are, or they are already excited about how fast their own Bambu printers are.

This is the first time Prusa Research hasn’t been the leader

There have always been fancier and more expensive 3D printers than the ones Prusa sells, but the more advanced printers always costs at least two or three times more. That isn’t the case any longer. Prusa Research is miles behind.

I don’t think Josef Prusa knows how to react. The Prusa MK2 and MK3 were easily the best printers under $3,000 for years. Then, suddenly you could buy a much more advanced Bambu X1C with a multimaterial unit for around the price of a fully assembled Prusa MK3S+ or Prusa MK4, and then you could buy the cut-down Bambu P1P for around the price of a Prusa MK3S+ or Prusa MK4 kit. He isn’t in the lead any longer. He isn’t even close.

Stable Diffusion

I do feel for him. He is in a tough spot, and he has at least 700 employees that are relying on him to keep his company in the lead.

They have been working on the Prusa XL for years, but I don’t believe Prusa has focused on speed. I don’t understand why they would design a Core-XY printer without focusing directly on speed. Did they really think the market didn’t want faster printers?

I like the Prusa MK4 as advertised

I don’t think $1,100 for a fully assembled Prusa MK4 is a bad value if it can print a Benchy in under 20 minutes. Especially knowing how little effort it has taken to keep my Prusa MK3S printing for two and a half years.

Bambu Lab can beat that 20 minutes with a printer that costs $750, and it can do it today. You don’t have to wait for a firmware update. We don’t even know when the firmware update for the Prusa MK4 is coming. We can print an 18-minute Benchy on a Bambu Lab printer right now.

The roughly equivalent competition costs less than the offerings from Prusa, and there are some amazing offerings from other companies at even lower price points today. My backup printer is a Sovol SV06 that I bought refurbished for $169. My only complaint about the Sovol so far is how loud it is. I don’t even need new printer slicer profiles. I send the exact same files to my Prusa and Sovol.

I haven’t had the Sovol long, but it has survived around 36 hours of printing a giant MINI 13 Bender. That doesn’t make it a good printer, but it does at least make me feel like it isn’t total junk.

Conclusion

This has been difficult to write. I always figure that if I can’t say anything nice, then I should just not say anything at all. The problem is that I am invested in my Prusa MK3S, the Prusa community, and the future product releases from Prusa Research.

Josef Prusa has spent years building trust with the community. It feels like he is tearing that trust down a little more every day. It is really bumming me out, and sometimes it even makes me angry. I would have been done with this blog two days ago if I didn’t keep getting grumpy, but then I would have missed out on including the Prusa MK4 filament debacle!

Bambu has yet to really earn my trust, but everyone I know who decided to buy one of their printers is having a great time. I am not someone constantly trying to chase the newest 3D printer. I use my printer to support my other hobbies. As long as my current machine is doing its job, I don’t have any reason to order a Bambu P1P or Bambu X1C.

Prusa hasn’t exactly lost all my trust, but this pattern is going in the wrong direction. Why would I pay $350 more for a Prusa MK4 than a bigger, faster Bambu P1P? It looks like the Prusa MK4 will only be begrudgingly open sourced, so the only reason left for me to buy my next 3D printer from Prusa Research is trust, and Prusa is spending that currency rather quickly right now.

Comments