Archicad 30: What have you done?

[Martin Jan Rosa]

Dear Graphisoft,

I've been an Archicad user for 25 years, and for all that time I've welcomed every new feature, big or small, with genuine enthusiasm. You even quoted my praise in the press release for the previous version.

With Archicad 30, though, I have nothing to praise.

Let's set aside the fact that the new release is little more than a handful of minor tweaks. The only thing I'd call noteworthy is the broader IFC 4.3 scope – which, frankly, is just you delivering what you promised for AC29 and quietly left half-done. Which makes me wonder: now that you've moved to subscriptions, why keep shipping "major" versions at all? Releases like this would sit perfectly well as regular updates to Archicad itself.

I'll also graciously overlook the fact that you shipped a version that crashes on the current macOS. I know you warn about it, but then again, Golden Gate has been in developers' hands for months – and you still couldn't get ready for it? I know, it's practically folklore by now: come December, it'll get quietly fixed somehow. Same goes for the icon you sneak into my dock after installation, linking to your new product – which isn't even available yet.

I won't dwell on the AI integration either. It still looks more like ticking off a mandatory buzzword than a feature anyone seriously intended to ship. It plays well in a board presentation; in practice, you reach for something else – something that does far more and keeps up with new AI capabilities far faster. The same goes for the MCP server. Nice to have, but any moderately AI-literate user can build their own in half an hour, and it will work better, because thanks to the community around Tapir it can do things you still can't offer. I sincerely hope common sense eventually prevails and you start taking AI seriously. Because unless Archicad becomes a real AI harness for the construction industry, I'd give it a few years before it goes the way of AutoCAD: a tool for professionals with an old-school mindset.

But what I truly cannot understand in Archicad 30 is something else entirely. Forgive me, but there's only one word for it – one you hear quite a lot these days: the enshittification of the user interface.

Working in Archicad 30 feels as if someone got a little carried away with the word "zen". The UI looks unfinished. The sensible response to mouse movement is gone. You intuitively lose track of where the workspace ends and the toolbars begin, where the buttons are, where the input fields are.

Honestly, it feels like you're making me work in Archicad blindfolded. Luckily, I know the interface well enough to guess where to click – otherwise I'd be half lost. Pity the poor new users.

So I'd kindly ask: could you perhaps take a little inspiration from Apple and their glass design – and take a similar step back, as they did? Maybe a simple setting that would de-zen the UI a bit and give it back some depth, structure and intuitiveness?

Many thanks

Martin Jan Rosa

[Eric Milberger]

The software we are using on our next project has up to 4 windows per monitor - 2 monitors equal 8 windows  in differing sizes

we have picked up our largest project in 60 years.  big decision to be made as I will add 4 more draftsmen

And at this moment Graphisoft is not a wise product as I see no direction and a blind eye to who the customers are and who are not.

A goo example Structural.  GRAPHISOFT only recommends one Structural client of theirs.  Portland I believe.  I can't visit there readily.  Great groop just too far away.  Why is that the only structural - maybe Graphisoft is not a good platform.  Or they don't understand how to make. their product full and strong.  Maybe they have 10,000 Engineers in Europe but in the U.S.not many outside residential projects.

[Bricklyne Clarence]

I wouldn't hold my breath expecting dual/multi-monitor capability in Windows Archicad in the next version, or any of the next several versions, if I were you.

It used to be a long-standing wish in the Wishlist section (it still is, to be fair), that was very popular and very popularly voted on by users - meaning that a lot of people want this as a feature development that Graphisoft should work on and integrate into a future version as a priority item.

And what did Graphisoft - the diligent, and 'very responsive to user-feedback' company that they are (/s) - do? Of course they did what any other company would do when faced with such a popular request. They placed it,...... "ON HOLD"....meaning, they were not working on implementing it anytime soon nor were there any plans to do so in the future.........For..... 'reasons'........ (The less said about those 'reason's and their explanation, the better for everyone).

And then when they were shamed when people noticed this, and then pointed out that most of the popularly voted on wishes were suddenly being placed "ON HOLD" (almost as if they didnt care at all about what users wanted or needed), while at the same time spending considerable resources and development time and talent introducing features that people LITERALLY never asked for (as in, don't even appear in the Wishlist section at all, much less have any votes)...they changed all those  "ON HOLD"'s to..... "Submitted".......which means.....what?...exactly? They gave some winded explanation which, true to their nature, didn't make any sense, but all of which boiled down to the idea that those wishes are still, for all intents and purposes,......."ON HOLD",......but just not being called so.

Just another sleight-of-hand rhetorical magic trick to try to fool people into thinking that anything has changed. A fresh coat of paint over an rickety piece of junk of a car.

It's of course embarrassing that a software of it's stature and position in the AEC industry and development world is still a single-monitor software in 20-freakin'-26. Even Notepad and MS Paint are dual-monitor, for crying out loud. 

And even dumber that this (IMHO), are the proposed or suggested workarounds like stretching  your archicad Window across two or more monitors,......which........what happens if I need to have another program's window also open (like a PDF-reader or an email message) at the same time in a different monitor while I'm working and I need to get information from it as I'm working? Or.....as you astutely pointed out....if one has monitors with different resolutions or refresh rates?

How is this even a "solution" that keeps getting suggested with a straight face or with any degree or seriousness, and not as a prank?

It's ridiculous!

Again,....in one man's humble opinion and with all respect where it's due.

If they never want to do it ever, then don't do it, and take it off the Wishlist section completely  and permanently, and put a open public memo informing new users that this will never be implemented in Archicad, with no reasons given. But don't insult our intelligence with silly workaround solutions that make no sense from a workflow point-of-view.

But I'm sure they'll get around to it.....eventually.......like,...maybe in Archicad v38 or v42,.... and then expect to be patted on the back and congratulated for finally giving Windows users, the functionality that Mac users have had for well over 15 years by that point, and which every other software in existence has had for the better part of this entire century since the early 2000's when multi-monitors first started to become a common everyday thing.

[Keny90]

I'm asking seriously, without any ill will: what has Graphisoft actually done in the name of development over the past year?

The AI is only in beta, and judging by the comments, it hasn't exactly won widespread approval, to put it mildly. I haven't tried it myself, since I don't meet the bar: I'm neither an Archicad Collaborate nor a Max subscriber. I don't really miss it either. With the help of Claude and Tapir, I've already written plenty of add-on tools that have replaced or improved certain functions, but that's not the topic here.

Apart from the AI add-on, the development package is practically negligible. I think we all agree on that, judging by the forum. It's been more or less like this since version 26, but after every release there were at least some positive voices in the posts. This time I couldn't find any…

Like the previous poster, I've been using Archicad since version 4.5. There were better and not-so-good years in terms of the improvements we received, but the direction was encouraging. You actually looked forward to the next year's release. Not every vintage of wine is excellent, after all. In recent years, though, that feeling has faded; they've been weak vintages. These days I'm more worried that they'll take away some feature I rely on. That's exactly what happened with the MEP add-on: it was gone from Archicad 27, and instead we got something unusable. I know few people used it, but I was one of those few. And the point isn't really MEP; it's the way Graphisoft handled it.

Anyway, it is what it is, and there's no point dwelling on it. We have to look ahead, and however deeply Archicad has become integrated into our company, if we have to let it go, we will. I've talked with several smaller, similar architectural firms with perpetual licenses about what they plan to do after version 30. Most of them said the same thing we're planning: this version will last us 2–3 years, especially if development continues at the pace of recent years. In the meantime, we'll try out and test other platforms, and we may switch to one, or we may stay…

I don't want to spin conspiracy theories, but I wouldn't be surprised if it's no coincidence that we got such negligible improvements this time. This release was aimed at perpetual license holders, and the bigger developments will probably be timed for mid-year or next year, to 'gently' motivate us to switch to a subscription. But that's just the personal opinion of a long-time, experienced user.

After this little detour, I still don't know: what has Graphisoft done in the name of development over the past year? If anyone knows anything, please enlighten me. Don't leave me in the dark. 😳

[Bricklyne Clarence]

---

@Keny90 wrote:

.........

Apart from the AI add-on, the development package is practically negligible. I think we all agree on that, judging by the forum. It's been more or less like this since version 26, but after every release there were at least some positive voices in the posts. This time I couldn't find any…

.....

---

I think you're being really generous there.

More than a few people have dubbed 'Archicad 30' as version "30" in name only, and that it is really Archicad v23.8 if we're summarizing it down to the real serious feature improvements that have been integrated since that version (or version 22, that had the lest serious major ARCHItectural tool upgrade and imrpovement in the Curtian Wall improvements. Design Options pushed it a version up to 23 but all other new "improvements" where Architcture users are concerned don't add up to warranting a full version number upgrade beyond that.)

I don't count any of the other major improvements they've had over the years as being seriously considered improvements from an "Architecture-first" perspective.

Things they've spent considerable development resources on like MEP addon/Designer, SAM tools, and tools that are supposedly still in some form of Beta or "Experimental" features like the 'AI toolset' or 'PBR visual display (despite them using these features to market new versions as if they're complete tools, when they know in reality  they're not).

And if they do what you're suggesting they might do and release a version 31 that's chock full of new features ( I doubt it) just to try and entice Perpetual license users  and holdouts to switch to Subscription, I think it will be too little and too late, and most have been making the decision to switch from Archicad to other solutions for the last few versions now.

Not to mention being a cynical way of approacing customer relations and dealing with customer expectations.

[RJK1457]

Like many companies that I have dealt with, as they grow and become more popular they push the little guy aside. I hope and pray that I am wrong, but over my 30 plus years in this business that seems to be the case. Auto Desk and Bently are no different. 

[RJK1457]

I would not get my hopes up, after seeing the marginal improvements since AC 26. My sales rep from Graphisoft USA just retired and I wonder why???? It's hard to sell nothing that's worth selling! ☹️ 

[jl_lt]

I'm still surprised that people is still surprised by what GS is doing (or better said, not doing).  But hey! that Nemetschek creator thing looks promising... For real estate developers that don't want to deal with architects of course

[Eric Milberger]

This is not a big deal in other softwear

[Alison_007]

It's sad to see what ArchiCAD is becoming. Since version 24-26 I haven't seen anything that justifies expecting an annual update. It's sad, but if it continues like this with the advent of AI, better options may emerge; I've seen many users developing alternative solutions to ArchiCAD or migrating.

[Alison_007]

And taking advantage of this opportunity, allow me to present a solution I'm developing that arose precisely from my frustrations and the many limitations I encountered during certain projects. I felt encouraged to develop this set of tools, and with the help of extensive research, AI, and basic C++ knowledge, I have achieved results that have already shown significant progress and are currently in the Beta/Release Candidate phase, available for Windows 10-11 x64 ArchiCAD 28, 29, and 30. 
In it, I integrate basic and advanced features that some have been asking for for years or decades. Such as a layer palette, Property Manager palette, direct selection and editing of zones, labels, text, advanced formatting options like making letters UPPERCASE, lowercase, merging text, automatically generating MEP network connections based on selection/ID, a tool to generate complete structures (beams, columns, slabs, zones, foundations, dimensions) based on walls or even based on DWG, and automatically generating linear dimensions/level dimensions in elevations or sections, image editing, and much, much more, as it's a Suite (Easy BIM Suite in a set of 8 integrated Add-Ons). I've attached some images for anyone interested; just access the link and test it, and perhaps support the project, as I intend to keep it free and continuously developing it. 
Link: Easy BIM Suite.

[Bricklyne Clarence]

What really bothers me is that someone - a single individual or maybe a small group of friends and hobbyists - can gather the motivation, the will and the little meager skills necessary to do what you've been able to do, and produce a suite of tools and addons that fill the very needs and areas we've been clamoring for GS to improve in the main program. And yet them with a full team of developers, vastly more resources (both within GS itself and from the wider Nemetschek ecosystem), constantly have to be pulled kicking and screaming to give us even the barest of minimum basic improvements in the program in any particular new release version.

And that's the few times they can spare some moments away from their precious side-projects and alternate side-gigs, chasing some other AEC-adjacent potential future customer-base like engineers, or planners and developers.

The advent of AI has been both a blessing and a curse, in that we are not slowly getting to the point (.....though not quite fully there yet) where base users can start having access to vibe-coded alternatives and options which weren't there before, and where the Autodesk's and GS's of the world could put out a bad or mediocre product and then say, "....if you don't like it, you can feel free to go kick rocks or use our rival's offering which we know you hate so much".

But on the other hand, it also entrenches them in this complacency where they feel they now don't have to prioritize improving any of the functions in the core program that your tools address since they can just point to it as a third party solution.

Kind of like how  GS themselves used the third-party addon availability for years to ignore improving large aspects of the program- and thus forcing users to have to spend even more money to get a 'complete' program over what they were already paying for license fees.

But we know there are some things that can't be vibe-coded to be improved in the program, and AI can take you only so far. And certain aspects of the program only GS themselves can actually improve since they're core to the program itself beyond what can be extended (like hardware related stuff like multi-monitor capability or simultaneous real-time window/tab updating, for example).

On a side-note, it's kind of ironic that the large software companies and monopolies like GS and Autodesk and others like Adobe the and Microsoft suites, all jumped on the AI train and bandwagon, gone all in on the hype and invested heavily in these systems - mostly as a means to get a new way to charge their customers for a new service that would be tied to their servers and their respective ecosystems.

And now we end up in a situation where AI and all the vibe-coded AI freeware opensource alternatives tot heir own software suddenly popping up are posing a real mortal threat to  many of them - with users all across the board already hating the subscription-only licensing model that they've all collectively adopted to make you pay money for something you don't own and for products with progressively decreasing and diminishing value in substance and a return on that investment.

Oh, the irony.

[Alison_007]

Hello!
It's truly sad to see the direction GS is taking, sad to see a program that I once praised highly in this state, and considering that practically everything they say is new is a kind of endless or incomplete beta version (Param-o, Python, GDL Editor, PBR mode, MEP Design, Mesh, Ramp and several other sectors/tools). 
I'm currently working on them alone, a job I started after a lot, and I mean a lot, of frustration, especially seeing what other software offers as basic features, let alone the advanced ones. 
So that's how I started (with the little knowledge of C++ I had and almost nothing of the ArchiCAD API) developing very simple and archaic tools to help me with simple tasks like creating zones in enclosed spaces independent of the floor and automated elevation cuts. 
But the development to this level it's at today only began in July of this year, and AI has been very useful and a major factor in achieving these results, not only quickly, but also by dedicating less time than before, which required a lot of research and review that I couldn't always solve. Now let's imagine this in the hands of truly qualified people? Seriously... Perfect, but... things are as they are. 
 
Finally, I would greatly appreciate it if you would try out the tools. Keeping in mind that this is a preview version before the official release, primarily aimed at ensuring they perform well in real-world testing and work scenarios, it's possible that one or two things might present some inconsistencies. 
I would greatly appreciate it if anyone could try it out and report any inconsistencies or malfunctions. I hope you really like it and find it very useful. 
 
Best Regards.

[Mats Knutsson]

In fairness AI has since summer totally turned everything upside down. Nightmare for developers since there is no way to choose the right path. I'd like to be a fly on the wall at various HQs right now. AC30 I've tested briefly but we user every odd version and will skip to 31. And yes...30 is another meh release. AC is still though a decent software but paying more and more for meh is not ideal.

[Martin Jan Rosa]

For developers it’s a nightmare, but for us users it’s a huge opportunity. I’ve already vibe-coded a number of add-ons and other tools myself, some of which I share and which are used by the local community in our market. And in many discussions I’ve always argued that this approach would never be able to produce a tool as complex as Archicad. Unfortunately, my experience with Archicad 30 inspired me last Wednesday to test what AI can currently do on exactly this task – and after less than a week, I have an app that is starting to come close to meeting my own needs (I should add that these are very limited compared to everything Archicad can do – but that’s precisely the point of this opportunity).

Honestly, I still have about 400 days left on my Archicad subscription, and now I simply have a completely new option that I’ll seriously consider when I decide in December 2027 whether to continue. There are already several tools I used for years and then replaced with my own, and just a month ago I wouldn’t have thought Archicad could be one of them. After all, I’m quite tied to it – both through more than two decades of user experience and through running the largest local community, where I’ve been helping users for years. But unfortunately, Archicad 30 was the culmination of my feeling that for many years now, Graphisoft hasn’t been developing in the interest of its users, but of its shareholders.

Fun fact: my video review of the new version, published today (the local community kind of expects one from me after every release…), starts with Archicad 30 crashing four times in a row at launch…

[strangeday]

Hi @Martin Jan Rosa , I have to say I had the exact same impression. I also believe that Archicad is the kind of software that’s truly difficult to replicate in the short term. However, for the past few months, I’ve been using AI tools to generate add-ons and objects that make my life much easier—and keep in mind, I don’t know how to write a single line of C++ or GDL code.

Honestly—and I don’t mean to be offensive to the Archicad developers—these tools are worth far more than the features introduced in Archicad versions 28 through 30.

If you look around, a number of BIM-related projects are popping up—like BIMCraft—which are still light-years away from replacing Archicad, but they are being developed at an astonishing speed (much like AI models, which become more powerful and capable of handling increasingly complex tasks every quarter). In my view, the floodgates are open; a lot is going to change over the next few years. We shall see.

P.S. Take a look at this: https://www.youtube.com/watch?v=t5xkc4mNXiY&t=173s. It was created by an Italian content creator/architect; it’s clearly a very simple example, but it was built entirely using Claude Code.

[Alison_007]

I truly believe that with AI, if many of these companies remain complacent, they could be in serious trouble. What used to take one or a handful of developers (today even someone with basic programming knowledge) months to develop, now takes less than half the time needed to achieve the same or even better results.

[cocoloco]

after reading everyone feedback, I am curious why do we actually need to download new Archicad version instead of incorporating these small changes into whatever we have installed? 

There is so much hussle with installing new versions,  transferring  ongoing projects and adjusting company settings , checking all imported correctly …. And all of that work on our part for what? A little updates? I have hard time justify downloading “new” Archicad version for the his little improvements 

[Pavel_N]

Agree with Martin. I have to say that the visual design of the interface is exceptionally poor. Input fields blend in—in fact, everything just blends together. I strongly urge you to fix this in the next update. It is a real failure and makes working with Archicad significantly more difficult.

[mthd]

Unfortunately our Archicad upgrades have slowed right down to a trickle these days. A bit like an internet connection that drops in and out and then slows right down. What might be causing this ? Many comments above blame greater investment in AI and this is true. As well as paying shareholders more money for less product. 

I think we all understand too, that once something is developed to a significant point it becomes harder and harder to improve it significantly. Take for example the development of the electric guitar, not much has changed since the early to mid 20th century. Apart from adding a few bells and whistles here and there, they have basically remained the same.

I still have access to CI tools because I bought perpetual licenses for some of them that I use quite frequently. Even though I am not on a collaborate subscription.

I haven’t got the time or knowledge to write add ons for AC as some are able to do. Is this the future, “Write your own add ons and don’t expect too much” ? Who knows but we will see what happens ?

To me it comes back to making sure that the basic functionality of the program is well serviced and not to be too invested in the bells and whistles that not everyone will use or even need.

[Funfun Yan]

I guess someone is working on their own Aurora with the help of AI.