It's been improving rapidly. The upcoming (imminently) 1.1 has a large amount of modern UI affordances, such as on-canvas gizmos that at times actually are easier to use than e.g. the Fusion ones. I'm a heavy Fusion user, but for me FreeCAD is nearly there now and the improvement over 1.0.x is massive.
There's a lot more to do, but my feeling is the project is taking UI/UX design much more seriously than it has in the past, with the ramp-up of an internal design-focused team etc. I get that feeling from reading the weekly progress updates and MR discussions.
I'm very optimistic for the future of FreeCAD personally. I think it's a great time to contribute if you are interested in making UI/UX better as well because there's much higher interest in that kind of work now. I think it's close to having its own Blender/KiCAD moment.
The UI has an awkward learning curve and some tools are weird, but it has become a rather solid CAD. Don't discount it in its current state, despite its warts.
Can't agree, I'm a complete newbie in CAD, and after I opened FreeCAD I didn't know what to do at all. Watched one youtube video covering all the basics and I can design whatever I want with confidence. Besides being free – it's very intuitive and great piece of software in my opinion
I disagree. If you switch to the Part Design workbench, it's basically the exact same workflow as SolidWorks. Draw a sketch, add constraints, extrude / revolve / fillet, etc.
Yes they have some unconventional names for certain operations, like "pad" instead of "extrude", and yes there's a confusingly-similarly-named "Part" workbench for doing CSG-style CAD, and yes it takes a bit of practice to get good at it. But it's not next to impossible.
I think it's fair. I use FreeCAD a lot, and the word I would use last to describe the UI is "discoverable". Ignoring whether a workflow is possible - whether the functionality exists at all - there's a whole lot of it that you have to Just Know, or equally as often Just Know That It's Not Actually Broken. The very fact that you started your comment with "If you <do X thing that you have to be told to do>" is precisely part of that.
I'm just a hobbyist with a 3D printer, but after watching a few tutorials it seems quite simple if you're only using the sketcher and part design 90% of the time.
Yeah I just end up using Fusion 360. But depending on a freebie for hobbyists that could be withdrawn any time it's a bit worrying. I wish there were good visual tools.
I don't like browser based and blender is too focused on animation (I'm more into 3D printing) so I haven't found a good FOSS alternative. FreeCAD isn't it anyway.
3dDune is actually pretty capable and simple for basic usecases.
Plasticity for free form hard surface CAD modeling.
If you need parametric CAD then the learning curve jumps significantly and FreeCAD nowdays actually makes sense as with a bit of practice and customization you get to great place.
FreeCAD feels like Blender few years before v2.80 - people outright dismissed it because it had a bit unconventional UX but underneath was already the extremely solid software that now dominates 3D polygon modeling. FreeCAD doesn't have that much of a momentum but i wouldn't be surprised if they became Blender of CAD over time.
Note it's only a freebie if you are on Windows/Mac, too. If you're on Linux, it works terribly on wine and you have to use the browser version, and then you need a $95/month subscription.
personally exited to check it out for real constructive-solid modeling, as opposed to emulating that workflow over OpenCascade's (fickle but otherwise lovely) BREP modeling (ie. edges & faces) via build123d (which has been great but is increasingly vibe-coded :/)
woah, thanks! seeing the scenes in the blog-post i realize i've ran into it before, but must not have observed the lineage, committed the project to memory, or realized it was so mature
there's even a parametric split-keyboard project (what i'm doing too)! the clearances and cutouts in julianschuler/concavum-customizer/.../keyboard/mod.rs[1] are so much like my static, single-file, build123d-based version in antlers/keyboard/.../main.rs[2] >u< (though i made the walls out of more layers, photo in README[3]). thx again for pointing me that way!
This thread has turned into a great resource! build123d has been my favourite conceptually so far (it's just Python) but vcad looks very clean too. I like the abuse of + and - in both of these for booleans.
I gave build123d a try about a year ago. I really wanted it to work, but it has a lot of issues, mainly in documentation. I'm going off memory here, and it's been a while, so maybe some of these have been fixed. One of the biggest issues is one of the fundamental classes (I want to say "Part") is not documented at all. And it's essentially the most important class. I tried enumerating all of the methods on the class, but didn't make much progress. Fillets were promising, but it seems once you've got a complex edge from a few operations it quits working, or at least did for my part. You're supposed to be able to do something like b=Box(10,10,10), then access the width as b.width, but all of those properties were always zero.
OpenSCAD supposedly supports Python now (https://pythonscad.org/), but I was not able to get it to work at all. I've fallen back to just OpenSCAD, even though it has limitations, at least I'm familiar with them. I'm mostly just waiting for improvements to anything that'll make it better than OpenSCAD.
i got really carried away making a contoured split-keyboard in build123d, and affirm there are some gaps in docs and functionality -- eg. it was quite tricky to get the center of an object in the right form (vector, position, or location), and often zero otherwise. i think a lot of it boils down to an OOP hierarchy with a lot of inherited or generit methods that aren't applicable or impl'ed. OpenCascade causes its own geom issues too, and by the end i was fighting build123d just to slap absolute-position--based hacks over inadvertantly invalid geometry. the dev team's interest in llm-generated code does not inspire confidence that it will become more reliable.
but i found-out up-thread that libfive, which i was esp. interested in for more reliable geom, has a Rust-based successor in Fidget -- and there's even a parametric split-keyboard that can be contrasted with my (static, hack-ier, but single-file) build123d iteration. exited to follow suit!
I've been "vibe coding" with OpenSCAD with good results! OpenSCAD will automatically detect changes in the current open file and reload it, so I can use VS Code (with the OpenSCAD extension) to vibe code with Claude, and watch the changes appear on the OpenSCAD screen
I've had better luck telling it to use CadQuery. Here's an example where I stumbled around a bit, but was successful in creating a cat food container (Sheba Perfect Portions) dispenser
They certainly have ambitions – the most recent changelog claims to add "Full PCB design pipeline: schematic capture, routing, DRC, Gerber export, and signal integrity simulation."
It also seems to have a physics engine, a slicer for 3D printing, an embroidery mode, and a entire ecosystem of math crates (https://tang.toys/).
Whether any of that works – or whether it's pure LLM slop – is less clear. I tried to import a trivial STEP file, and it crashed my browser tab [1]. Every commit is co-authored by Claude.
So far, he’s shown incredible productivity (with Claude Code). I integrated his vcad into my toy project here, and it worked on the first try, which is quite impressive for such a young project:
https://github.com/darwin/supex/tree/dev
"Electronic design automation (EDA), also referred to as electronic computer-aided design (ECAD),[1] is a category of software tools for designing electronic systems such as integrated circuits and printed circuit boards."
Okay, but unless you choose to download the Windows executable, compiling from source is very difficult. Many people won't accept the Snap option on an otherwise open-source platform.
This project improves on SolveSpace, but it does this by requiring dozens of mutually conflicting libraries. I create CAD videos, but for my students I decided against this project after seeing how difficult it was to compile.
A FlatPak installer might help with this installation issue.
Again, the Windows executable gets around these issues, for people still willing to put up with Windows.
Second reply -- if anyone wants to run Dune3D, flatpak or compiled, they must set this flag in advance:
export GDK_DEBUG="gl-prefer-gl"
I discovered this while trying out the compiled version (it's essential for the program to run at all), and for some reason I thought the FlatPak install would have done away with this oddity.
Again, because my students aren't necessarily techies, this kind of hacking shouldn't exist in a program released to mere mortals.
But thanks again for alerting me to this release version.
Dune3d comes off like a reskin of SolveSpace. SolveSpace is pretty awesome, so that's not a knock, per se. I'll leave it to somebody with more experience to fill in what value Dune3D adds beyond SolveSpace.
I found the UI far more approachable in Dune 3D than any other 3D CAD program I've tried and as the readme notes, Dune 3D imports STEP files and does fillets/chamfers which SolveSpace does not (in the current version)
If someone knows of a general introduction to 3D CAD which focuses on vendor-neutral descriptions of terminology and concepts, I'd be very interested --- I've done the tutorial for Dune 3D twice now (which is farther than I've gotten in any other 3D CAD tool), but keep getting hung up on subtleties/specifics which I have trouble describing for want of the correct terminology/understanding:
I've had far more success following along video tutorials than written ones. Most written tutorials (as you've pointed out) miss far too much detail. Watching someone do it and copying along teaches all the menu navigation stuff implicitly.
I've successfully learned quite a few EDA and 3D CAD tools that way. It's also effectively the way it's taught in a classroom - the teacher shows you and you copy.
I've been trying that, but I've been having a hard time catching the nuances of terminology and minor differences in UI --- hopefully articles such as this will help to make Dune 3D more popular and more videos will show up.
> If someone knows of a general introduction to 3D CAD which focuses on vendor-neutral descriptions of terminology and concepts, I'd be very interested
I'm not sure there is such a thing. Maybe a drafting book would be a vendor neutral description of spec'ing a 3D manufacturable object on a 2D surface.
Otherwise the way the software accomplishes the task is kinda the point which will necessarily have methodology specific terms in it.
The documentation addresses that[0]. Basically, Dune 3D uses solvespace's solver, but it can do fillets and chamfers, and has a slightly more approachable user interface.
The main things Dune3D has over solvespace are 1 - the ability to include STEP files in assemblies, and 2 - the ability to put a chamfer or fillet on the top/bottom of an extrusion. I don't think it has the ability to select arbitrary edges and apply such features (I could be wrong).
Solvespace 3.2 is essentially done (try the latest edge build), we just need to get github to produce the correct artifacts (not sure what broke). After that we will be adding named parameters to be used in sketches, and I'll try to get back to the chamfer/fillet problem and NURBS issues.
> I don't think it has the ability to select arbitrary edges and apply such features (I could be wrong).
That's actually how it works. Chamfers and fillets are handled by OpenCascade, so they can be applied to arbitrary edges. They only exist on the solid model though, so they are not integrated with the solver.
I'd be curious to see how native chamfers/fillets in solvespace would work, I imagine some amazing things could be done if they'd be part of the solver.
Named parameters will be _huge_ and if it could be considered, I'd dearly love to see a human-oriented input/export format where an individual could edit a text file, import it to get an interactive/graphical representation, then export to the same format for subsequent editing --- if nothing else, a format such as this would allow easy versioned storage of files.
SolveSpace is a PITA in that regard. You also need to re-learn most terms that are common in other CAD software. It's a typical OSS thing.
Why care about professional users that have years of learning invested into an ecosystem of professional CAD software (including terminology)? Because these people will get you the most valuable feedback if you can get them to even play with your OSS CAD thingy.
AI has now balanced the scores here. Someone with decent CAD experience can now instruct a model to build something useful.
Based on lots of good libs out there that solve the basics. I.e. concentrate on UI/UX to build something better.
It's like Lego, hands-free. You have all the blocks and you have someone who knows how/helps you combine them.
If you have good taste, you can get nice results without understanding how the Legos where made or how to even combine them.
I agree that it is very different from other CAD software, but that is also something I like. When I started using SolveSpace for some easy models, I was a bit lost. But then it clicked, and I really enjoyed the different approach. It is not my main CAD software (build123d is), but I really appreciate the workflow.
I don't think that every oss should try the copy what already exists. The best is when new approaches are tried. The same happened to me when I started using tilling window managers. "Professional" Operating systems don't have that, but I am sure that if more people would try them, many would realize that the workflow fits them better. So, my point is that there is no single best solution in terms of user interface or interaction with a program, and the fact that many people explore and share different approaches with their open source software is something I really appreciate.
Also check out modelrift.com which is based on openscad foundation. See the dynamic customizer which allows to edit any model parameter, re-render and get .stl: https://modelrift.com/models/customizable-liquid-funnel - it works _completely_ in your browser by using WASM
Would love to see a quick video demo showcasing the features, look and feel of the software. The same team made horizon eda, and I wasn't able to find videos on youtube about it that were newer than 2022, so I never gave it a shot either.
I have used this for throwing together some models for 3D printing. I've found it very intuitive, though I'm not sure how ergonomic it would be for complex assemblies.
I really like the space-key based command access and default shortcuts for all the commands.
So this is a smudge of like 4 projects? Huh. Definitely interested, but I wonder about the longevity of the system. That's one thing about the code cad systems I like: it's pretty easy to port code from one to another.
In case anyone is wondering, Dune3D as a flatpak is about 33mb. FreeCAD is 354mb. I enjoy having simple solutions that get simple things done. Will definitely give Dune3D a try.
Maybe Shapr3D? (for folks who are willing to pay monthly) or Moment of Inspiration 3D (commercial, perpetual license, by the former lead developer of Rhino 3D https://www.youtube.com/watch?v=H9k4ZbcnW-Y )
It's by the team that did Horizon EDA, which is very much a "craftsmanship-first" effort. and they've been working on this for some time now. My guess is hardly any.
Is it a team really? Most commits seem to come from one user "carrotindustries". I am really interested in an Open Source CAD application with good UX, this one looks great. But I don't want to spend too much time on an application maintained mostly by a single developer. The risk of it being abandoned is too high.
(I seem to be cast in the role of FreeCAD advocate on HN these days, but here goes!)
For years I agreed with you - I tried FreeCAD multiple times, different versions, always sucked.
Then I watched this video [0] and discovered that v1.1 is different - and that it's good enough for solid reliable hobby usage. It's still a touch frustrating in a few areas (text, for example) but I've now switched over to it completely.
I've forced myself to git gud with FreeCAD. It's better. Way better than it used to be. It's also still a very complex and user unfriendly application with a long road ahead of it.
You can make it work. You can also save yourself a lot of headache by using other CAD tools. Personally I value "Freedom" so I will continue to use it despite the difficulties but that may not be the right path for others.
I agree. freeCAD has become a tool that I just use without thinking about it. Earlier versions always made me question my choice and try out other software.
I really, really want that to be true, but my experience trying to adopt it has been really painful.
Even selecting things in the UI has sucked. I went in and increased the selection radius or whatever, that helped. But really, should I need to do this as a new user?
Getting the constraints to behave is like pulling teeth.
It also kind of sucks that you have to have really sparse sketches that only contain one closed figure. I gather you can create a "master sketch" and selectively project geometry into other sketches. But the last few times I've tried the app, I haven't gotten far enough into my sketches before rage quitting to validate the technique.
Right now I am back F360 with their hobby license wanting to escape their regular messing with the terms and conditions.
> Even selecting things in the UI has sucked. I went in and increased the selection radius or whatever, that helped. But really, should I need to do this as a new user?
Agree - selection isn’t broken, but it’s definitely sometimes frustrating and as it’s such a common function, absolutely should be as close to perfect as possible. I think it’s partly that the visual indication of what you’re hovering over and would be selected is too subtle, and also I’ve found (on Mac; I’ve not confirmed on other OSs) that it’s not selecting what’s at the exact tip of the pointer, but is rather selecting a couple of pixels away.
> Getting the constraints to behave is like pulling teeth.
Huh, once I’ve actually selected correctly, I find the constraints are fine - say, 95% as good as Solidworks.
> It also kind of sucks that you have to have really sparse sketches that only contain one closed figure.
Can you explain what you mean by this? Do you mean you can’t have a sketch with (to take a very simple example) a circle inside a circle, or two unrelated circles, or something else?
There are situations I can think of where selection does seem broken by design. It's fairly easy to get into a situation in the 3d view where you want to select a vertex but because of the draw order it's very hard to find an orientation of the model that lets you put it "in front". So you spend ages selecting the lines around it, spinning the model, trying again from all sorts of angles. Heaven help you if you're trying to select a bunch of points that have this problem, it's frustrating as hell. The second is in sketches, where the constraint icons aren't selectable when they're grouped but will block the selection of a component underneath them anyway. That's just obnoxious. I think in both cases the UI is working as designed, but it makes for an unusable outcome.
Oh, and if the selection point isn't at the pointer point? That's just a bug, and needs to be fixed. I can't see any defending that.
I’ve tried FreeCAD multiple times, but I’m just too used to Rhino 3D. Unfortunately, it’s rather expensive. Anyone need a slightly used, low-mileage, one owner soul?
FWIW, I've been kind of kicking myself for being bamboozled into buying a Plasticity Studio license --- really, really wish I'd instead bought Moment of Inspiration 3D (which is by the lead developer of Rhino 3D).
Curious what folks think of MOI3D (and if there is any other similar CAD tool other than maybe Shapr3D).
To echo others' comments: FreeCAD has improved significantly since v1.0, so I'm hoping this attracts quality & stability-minded develeopers, and a frequent release cadence.
> Its not often you see 'fillets and chamfers' are tip-line features in the readme for CAD packages.
Well the readme states the following:
Solvespace on the other hand gets the workflow part right, but falls short by not importing STEP and the geometry kernel not supporting chamfers and fillets.
This is honestly the first thing I look for with anything new claiming "CAD".
Roughly every other week there is a new "The (programmable) CAD that fixes everything!" post on the front page, just for me to open them up excitedly and noticing that they use a mesh kernel and will thus never be able to provide fillets and chamfers painlessly (for the user). All while they are absolutely essential for a lot of designs, especially in 3D-printing, a well-placed fillet/chamfer can make the difference between an object that breaks upon looking at it funny and one that can bear significant load.
Yeah, it's a tough row to hoe --- I've been handling it quite differently in my own project, taking "Design for Manufacturing" to the ultimate conclusion and requiring that the part be designed by actually applying tooling:
Needs a full-on re-write to make the G-code export (and import) work well, but lately I've been focused on exporting to DXFs w/ colour/layer tagging which was just added as an import feature in the CAD tool my side-gig employer does.
G2 fillets are the next frontier. Even Fusion doesn't seem to handle them well. I mean they can be created without much drama, but geometry derived from them is very likely to fail with no real reason.
Hull and Minkowski operations sort of allow one to do this sort of thing, but add to the mathematical complexity in what is probably the worst-possible way....
Some other open source 3D CAD tools:
Code-based
- CadQuery - https://github.com/CadQuery/cadquery/
- build123d - https://github.com/gumyr/build123d
- OpenSCAD - https://openscad.cloud/openscad/
GUI (browser-based)
- Cadmium (abandoned, cool idea) - https://mattferraro.dev/posts/cadmium
FreeCAD: https://freecad.org
I would not suggest anyone use FreeCAD. The UX is the worst I've seen for any software. Finding any functionality is next to impossible.
It's been improving rapidly. The upcoming (imminently) 1.1 has a large amount of modern UI affordances, such as on-canvas gizmos that at times actually are easier to use than e.g. the Fusion ones. I'm a heavy Fusion user, but for me FreeCAD is nearly there now and the improvement over 1.0.x is massive.
There's a lot more to do, but my feeling is the project is taking UI/UX design much more seriously than it has in the past, with the ramp-up of an internal design-focused team etc. I get that feeling from reading the weekly progress updates and MR discussions.
I'm very optimistic for the future of FreeCAD personally. I think it's a great time to contribute if you are interested in making UI/UX better as well because there's much higher interest in that kind of work now. I think it's close to having its own Blender/KiCAD moment.
The UI has an awkward learning curve and some tools are weird, but it has become a rather solid CAD. Don't discount it in its current state, despite its warts.
Can't agree, I'm a complete newbie in CAD, and after I opened FreeCAD I didn't know what to do at all. Watched one youtube video covering all the basics and I can design whatever I want with confidence. Besides being free – it's very intuitive and great piece of software in my opinion
Which video did you watch?
this one from CAD CAM lessons: https://www.youtube.com/watch?v=jULWgMV9_TM
This guys videos are up to date and always great too, if anyone missed them https://www.youtube.com/@MangoJellySolutions
That's pretty rough criticism considering FreeCAD how great work FreeCAD team has been doing on UX recently.
Also all of the software mentioned above has very little UI/UX and also ambition compared to FreeCAD.
I disagree. If you switch to the Part Design workbench, it's basically the exact same workflow as SolidWorks. Draw a sketch, add constraints, extrude / revolve / fillet, etc.
Yes they have some unconventional names for certain operations, like "pad" instead of "extrude", and yes there's a confusingly-similarly-named "Part" workbench for doing CSG-style CAD, and yes it takes a bit of practice to get good at it. But it's not next to impossible.
I think it's fair. I use FreeCAD a lot, and the word I would use last to describe the UI is "discoverable". Ignoring whether a workflow is possible - whether the functionality exists at all - there's a whole lot of it that you have to Just Know, or equally as often Just Know That It's Not Actually Broken. The very fact that you started your comment with "If you <do X thing that you have to be told to do>" is precisely part of that.
I'm just a hobbyist with a 3D printer, but after watching a few tutorials it seems quite simple if you're only using the sketcher and part design 90% of the time.
Yeah I just end up using Fusion 360. But depending on a freebie for hobbyists that could be withdrawn any time it's a bit worrying. I wish there were good visual tools.
I don't like browser based and blender is too focused on animation (I'm more into 3D printing) so I haven't found a good FOSS alternative. FreeCAD isn't it anyway.
3dDune is actually pretty capable and simple for basic usecases.
Plasticity for free form hard surface CAD modeling.
If you need parametric CAD then the learning curve jumps significantly and FreeCAD nowdays actually makes sense as with a bit of practice and customization you get to great place.
FreeCAD feels like Blender few years before v2.80 - people outright dismissed it because it had a bit unconventional UX but underneath was already the extremely solid software that now dominates 3D polygon modeling. FreeCAD doesn't have that much of a momentum but i wouldn't be surprised if they became Blender of CAD over time.
I don't really do simple designs, and making them parametric would be really difficult. I'm just more visually oriented than mathematical.
But yeah FreeCAD I just can't get my head around it right now.
Try https://www.plasticity.xyz/
Hmm interesting but no free option for hobbyists (just 30 days). That's a bit of a dealbreaker for me. Looks like a good tool though.
The price is a lot lower than Fusion but still too high for hobby use IMO.
Autodesk making Fusion worse in any way could be the catalyst for FreeCAD to realize its full potential
There is quite a bit more competition though. For example OnShape is imho sweetspot and it's also kinda free.
Note it's only a freebie if you are on Windows/Mac, too. If you're on Linux, it works terribly on wine and you have to use the browser version, and then you need a $95/month subscription.
Yeah I have a Windows gaming box which I use so I use it on that. I would not use a 3D design tool in the browser anyway.
It depends; it’s not bad with the right plug-ins for designing a building, but God forbid you ever have to model a threaded hole.
Cadova - https://github.com/tomasf/Cadova
SolveSpace - https://github.com/solvespace/solvespace
libfive - https://libfive.com
guile scheme, bindings in Rust and Python
personally exited to check it out for real constructive-solid modeling, as opposed to emulating that workflow over OpenCascade's (fickle but otherwise lovely) BREP modeling (ie. edges & faces) via build123d (which has been great but is increasingly vibe-coded :/)
discussed previously: https://news.ycombinator.com/item?id=12319406
a real constructive solid library (as opposed to emulation via modeling), with
Libfive has been superseded long ago by fidget (same author, Rust not C++).
https://www.mattkeeter.com/projects/fidget/
https://github.com/mkeeter/fidget
woah, thanks! seeing the scenes in the blog-post i realize i've ran into it before, but must not have observed the lineage, committed the project to memory, or realized it was so mature
there's even a parametric split-keyboard project (what i'm doing too)! the clearances and cutouts in julianschuler/concavum-customizer/.../keyboard/mod.rs[1] are so much like my static, single-file, build123d-based version in antlers/keyboard/.../main.rs[2] >u< (though i made the walls out of more layers, photo in README[3]). thx again for pointing me that way!
1: https://github.com/julianschuler/concavum-customizer/blob/ma...
2: https://codeberg.org/antlers/keyboard/src/branch/main/src/ca...
3: https://codeberg.org/antlers/keyboard
JSCAD (formerly OpenJSCAD) https://openjscad.xyz/
Awesome because you can build a model, expose the parameters, and allow web users to generate a model to fit their parameters.
Truck (Rust CAD Kernel) - https://github.com/ricosjp/truck
This thread has turned into a great resource! build123d has been my favourite conceptually so far (it's just Python) but vcad looks very clean too. I like the abuse of + and - in both of these for booleans.
I gave build123d a try about a year ago. I really wanted it to work, but it has a lot of issues, mainly in documentation. I'm going off memory here, and it's been a while, so maybe some of these have been fixed. One of the biggest issues is one of the fundamental classes (I want to say "Part") is not documented at all. And it's essentially the most important class. I tried enumerating all of the methods on the class, but didn't make much progress. Fillets were promising, but it seems once you've got a complex edge from a few operations it quits working, or at least did for my part. You're supposed to be able to do something like b=Box(10,10,10), then access the width as b.width, but all of those properties were always zero.
OpenSCAD supposedly supports Python now (https://pythonscad.org/), but I was not able to get it to work at all. I've fallen back to just OpenSCAD, even though it has limitations, at least I'm familiar with them. I'm mostly just waiting for improvements to anything that'll make it better than OpenSCAD.
i got really carried away making a contoured split-keyboard in build123d, and affirm there are some gaps in docs and functionality -- eg. it was quite tricky to get the center of an object in the right form (vector, position, or location), and often zero otherwise. i think a lot of it boils down to an OOP hierarchy with a lot of inherited or generit methods that aren't applicable or impl'ed. OpenCascade causes its own geom issues too, and by the end i was fighting build123d just to slap absolute-position--based hacks over inadvertantly invalid geometry. the dev team's interest in llm-generated code does not inspire confidence that it will become more reliable.
but i found-out up-thread that libfive, which i was esp. interested in for more reliable geom, has a Rust-based successor in Fidget -- and there's even a parametric split-keyboard that can be contrasted with my (static, hack-ier, but single-file) build123d iteration. exited to follow suit!
I've been "vibe coding" with OpenSCAD with good results! OpenSCAD will automatically detect changes in the current open file and reload it, so I can use VS Code (with the OpenSCAD extension) to vibe code with Claude, and watch the changes appear on the OpenSCAD screen
Got any example results/chat sessions? I've had little luck with LLMs for 3d modeling
I've had better luck telling it to use CadQuery. Here's an example where I stumbled around a bit, but was successful in creating a cat food container (Sheba Perfect Portions) dispenser
https://claude.ai/share/ebce7c8e-4e5a-42ec-8ee9-cf066f68858f
You might have an easier time doing that with claude code.
Possibly, but for reasonably quick tasks it's nice to get previews or downloadable artifacts in the UI (I use the Claude Desktop app)
try modelrift.com ? Llms are bad at spatial understanding but you can steer them using annotated screenshots ("annotation mode" of modelrift) and get something working: https://modelrift.com/models/cable-conceal-box-with-wall-mou...
https://modelrift.com/models/customizable-liquid-funnel
Brl-cad https://brlcad.org/
vcad was started in 2026 and has great ambitions: https://docs.vcad.io
They certainly have ambitions – the most recent changelog claims to add "Full PCB design pipeline: schematic capture, routing, DRC, Gerber export, and signal integrity simulation."
It also seems to have a physics engine, a slicer for 3D printing, an embroidery mode, and a entire ecosystem of math crates (https://tang.toys/).
Whether any of that works – or whether it's pure LLM slop – is less clear. I tried to import a trivial STEP file, and it crashed my browser tab [1]. Every commit is co-authored by Claude.
[1] https://github.com/ecto/vcad/issues/7
Thanks for the bug report! I'll have some time later this week to look into it. Just had a baby :)
Congrats!
(We had one back in December; you’re in for a fun ride!)
...and don’t forget Loon Lang — it’s a gem: https://loonlang.com
By the way, “they” is actually just one person: Cam Pedersen — https://campedersen.com
So far, he’s shown incredible productivity (with Claude Code). I integrated his vcad into my toy project here, and it worked on the first try, which is quite impressive for such a young project: https://github.com/darwin/supex/tree/dev
Definitely keep an eye on him.
<3
KiCad - https://www.kicad.org/
that's for PCB design, it's not a 3D CAD tool
The Wikipedia template for CAD software
> https://en.wikipedia.org/wiki/Template:CAD_software
distinguishes between 5 types of CAD software:
1. Mechanical
2. Architectural (AEC)
3. Electrical
4. Optical
5. Garment
Thus: KiCad is clearly a CAD application, though not of the Mechanical, but of the Electrical category (and is listed in the linked template as such).
With all due respect to Wikipedia
https://arshon.com/blog/eda-vs-cad-decoding-the-basics-of-de...
And if we're appealing to Wikipedia as an (inconsistent) authority, https://en.wikipedia.org/wiki/Electronic_design_automation
"Electronic design automation (EDA), also referred to as electronic computer-aided design (ECAD),[1] is a category of software tools for designing electronic systems such as integrated circuits and printed circuit boards."
And
https://en.wikipedia.org/wiki/KiCad
"KiCad (/ˈkiːˌkæd/ KEE-kad[7]) is a free software suite for electronic design automation (EDA)."
But KiCad did put CAD in their own name so ¯\_(ツ)_/¯
;-;
Okay, but unless you choose to download the Windows executable, compiling from source is very difficult. Many people won't accept the Snap option on an otherwise open-source platform.
This project improves on SolveSpace, but it does this by requiring dozens of mutually conflicting libraries. I create CAD videos, but for my students I decided against this project after seeing how difficult it was to compile.
A FlatPak installer might help with this installation issue.
Again, the Windows executable gets around these issues, for people still willing to put up with Windows.
There is a flatpak https://flathub.org/en/apps/org.dune3d.dune3d
Second reply -- if anyone wants to run Dune3D, flatpak or compiled, they must set this flag in advance:
export GDK_DEBUG="gl-prefer-gl"
I discovered this while trying out the compiled version (it's essential for the program to run at all), and for some reason I thought the FlatPak install would have done away with this oddity.
Again, because my students aren't necessarily techies, this kind of hacking shouldn't exist in a program released to mere mortals.
But thanks again for alerting me to this release version.
Thanks -- I missed this. If it pans out, it might get me to shift away from SolveSpace, which has a few perpetually annoying quirks.
took me two commands to get binary on osx. (i had brew already)
brew install ... and ./scripts/build_macos.sh
For windows instructions look equally trivial.
Even an AppImage would be great.
Without context, it sounds like an FPS adaptation of a Dune game.
Yep, or a 3D adaptation of an RTS game.
3D adaptation of the 1992 Cryo Interactive adventure game.
The original CD version already had 3D.
https://www.youtube.com/watch?v=aQ2U8_3EE9k&t=1000
Thanks for unlocked memories!
Dune3d comes off like a reskin of SolveSpace. SolveSpace is pretty awesome, so that's not a knock, per se. I'll leave it to somebody with more experience to fill in what value Dune3D adds beyond SolveSpace.
I found the UI far more approachable in Dune 3D than any other 3D CAD program I've tried and as the readme notes, Dune 3D imports STEP files and does fillets/chamfers which SolveSpace does not (in the current version)
There was a recent video on it:
https://www.youtube.com/watch?v=T1VNpC0nwF4
If someone knows of a general introduction to 3D CAD which focuses on vendor-neutral descriptions of terminology and concepts, I'd be very interested --- I've done the tutorial for Dune 3D twice now (which is farther than I've gotten in any other 3D CAD tool), but keep getting hung up on subtleties/specifics which I have trouble describing for want of the correct terminology/understanding:
https://github.com/dune3d/dune3d/discussions/118
When I tried to write up the usage of a far simpler program, one of the things which I tried to do was define all terminology as it was brought up:
https://willadams.gitbook.io/design-into-3d/2d-drawing
are there any tutorials for 3D CAD which attempt definitions along the way in this fashion?
I've had far more success following along video tutorials than written ones. Most written tutorials (as you've pointed out) miss far too much detail. Watching someone do it and copying along teaches all the menu navigation stuff implicitly.
I've successfully learned quite a few EDA and 3D CAD tools that way. It's also effectively the way it's taught in a classroom - the teacher shows you and you copy.
I've been trying that, but I've been having a hard time catching the nuances of terminology and minor differences in UI --- hopefully articles such as this will help to make Dune 3D more popular and more videos will show up.
Another is:
https://www.youtube.com/watch?v=nsOHTD1RYBY
I wish the person doing:
https://ocw.mit.edu/courses/res-16-002-how-to-cad-almost-any...
would revisit that course w/ Dune 3D.
> If someone knows of a general introduction to 3D CAD which focuses on vendor-neutral descriptions of terminology and concepts, I'd be very interested
I'm not sure there is such a thing. Maybe a drafting book would be a vendor neutral description of spec'ing a 3D manufacturable object on a 2D surface.
Otherwise the way the software accomplishes the task is kinda the point which will necessarily have methodology specific terms in it.
But elsethread there's a person complaining that industry standard terminology is _not_ used in Solvespace....
The documentation addresses that[0]. Basically, Dune 3D uses solvespace's solver, but it can do fillets and chamfers, and has a slightly more approachable user interface.
0: https://docs.dune3d.org/en/latest/why-another-3d-cad.html
The main things Dune3D has over solvespace are 1 - the ability to include STEP files in assemblies, and 2 - the ability to put a chamfer or fillet on the top/bottom of an extrusion. I don't think it has the ability to select arbitrary edges and apply such features (I could be wrong).
Solvespace 3.2 is essentially done (try the latest edge build), we just need to get github to produce the correct artifacts (not sure what broke). After that we will be adding named parameters to be used in sketches, and I'll try to get back to the chamfer/fillet problem and NURBS issues.
> I don't think it has the ability to select arbitrary edges and apply such features (I could be wrong).
That's actually how it works. Chamfers and fillets are handled by OpenCascade, so they can be applied to arbitrary edges. They only exist on the solid model though, so they are not integrated with the solver.
I'd be curious to see how native chamfers/fillets in solvespace would work, I imagine some amazing things could be done if they'd be part of the solver.
Named parameters will be _huge_ and if it could be considered, I'd dearly love to see a human-oriented input/export format where an individual could edit a text file, import it to get an interactive/graphical representation, then export to the same format for subsequent editing --- if nothing else, a format such as this would allow easy versioned storage of files.
Usability?
SolveSpace is a PITA in that regard. You also need to re-learn most terms that are common in other CAD software. It's a typical OSS thing.
Why care about professional users that have years of learning invested into an ecosystem of professional CAD software (including terminology)? Because these people will get you the most valuable feedback if you can get them to even play with your OSS CAD thingy.
AI has now balanced the scores here. Someone with decent CAD experience can now instruct a model to build something useful.
Based on lots of good libs out there that solve the basics. I.e. concentrate on UI/UX to build something better.
It's like Lego, hands-free. You have all the blocks and you have someone who knows how/helps you combine them.
If you have good taste, you can get nice results without understanding how the Legos where made or how to even combine them.
I agree that it is very different from other CAD software, but that is also something I like. When I started using SolveSpace for some easy models, I was a bit lost. But then it clicked, and I really enjoyed the different approach. It is not my main CAD software (build123d is), but I really appreciate the workflow.
I don't think that every oss should try the copy what already exists. The best is when new approaches are tried. The same happened to me when I started using tilling window managers. "Professional" Operating systems don't have that, but I am sure that if more people would try them, many would realize that the workflow fits them better. So, my point is that there is no single best solution in terms of user interface or interaction with a program, and the fact that many people explore and share different approaches with their open source software is something I really appreciate.
Is there a good text (or video) which discusses the basics of 3D CAD and all the attendant terminology and concepts?
FWIW, there has been some previous discussion here:
1.2 https://news.ycombinator.com/item?id=41975958
1.1 https://news.ycombinator.com/item?id=40228068
1.0 https://news.ycombinator.com/item?id=37979758
(unfortunately, no discussion for 1.3, so I guess this stands in for 1.4)
Also check out modelrift.com which is based on openscad foundation. See the dynamic customizer which allows to edit any model parameter, re-render and get .stl: https://modelrift.com/models/customizable-liquid-funnel - it works _completely_ in your browser by using WASM
Would love to see a quick video demo showcasing the features, look and feel of the software. The same team made horizon eda, and I wasn't able to find videos on youtube about it that were newer than 2022, so I never gave it a shot either.
Just watched this the other day: https://youtu.be/T1VNpC0nwF4?si=T9XDd2_2sY0BaBFZ
I have used this for throwing together some models for 3D printing. I've found it very intuitive, though I'm not sure how ergonomic it would be for complex assemblies.
I really like the space-key based command access and default shortcuts for all the commands.
You can play Dune for DOS here: https://dos.zone/dune-cryo/
So this is a smudge of like 4 projects? Huh. Definitely interested, but I wonder about the longevity of the system. That's one thing about the code cad systems I like: it's pretty easy to port code from one to another.
Another simple GTK4 app for the ecosystem, nice.
In case anyone is wondering, Dune3D as a flatpak is about 33mb. FreeCAD is 354mb. I enjoy having simple solutions that get simple things done. Will definitely give Dune3D a try.
is there something like sketchup make that isn't stuck in 2017?
Maybe Shapr3D? (for folks who are willing to pay monthly) or Moment of Inspiration 3D (commercial, perpetual license, by the former lead developer of Rhino 3D https://www.youtube.com/watch?v=H9k4ZbcnW-Y )
How much vibe coding is involved in this?
It's by the team that did Horizon EDA, which is very much a "craftsmanship-first" effort. and they've been working on this for some time now. My guess is hardly any.
Is it a team really? Most commits seem to come from one user "carrotindustries". I am really interested in an Open Source CAD application with good UX, this one looks great. But I don't want to spend too much time on an application maintained mostly by a single developer. The risk of it being abandoned is too high.
As opposed to what?
Autodesk cogitating on what feature they can next do a rug-pull on?
The source is available, worst-case scenario is one puts together the money to hire a developer to fix or improve at need.
None, at least not deliberately. Previous attempts at vibe-coded contributions have been politely rejected.
Source: I am one of the contributors.
FreeCAD is the worst. Thanks for building this!
(I seem to be cast in the role of FreeCAD advocate on HN these days, but here goes!)
For years I agreed with you - I tried FreeCAD multiple times, different versions, always sucked.
Then I watched this video [0] and discovered that v1.1 is different - and that it's good enough for solid reliable hobby usage. It's still a touch frustrating in a few areas (text, for example) but I've now switched over to it completely.
[0] https://youtu.be/VEfNRST_3x8
I've forced myself to git gud with FreeCAD. It's better. Way better than it used to be. It's also still a very complex and user unfriendly application with a long road ahead of it.
You can make it work. You can also save yourself a lot of headache by using other CAD tools. Personally I value "Freedom" so I will continue to use it despite the difficulties but that may not be the right path for others.
I agree. freeCAD has become a tool that I just use without thinking about it. Earlier versions always made me question my choice and try out other software.
I really, really want that to be true, but my experience trying to adopt it has been really painful.
Even selecting things in the UI has sucked. I went in and increased the selection radius or whatever, that helped. But really, should I need to do this as a new user?
Getting the constraints to behave is like pulling teeth.
It also kind of sucks that you have to have really sparse sketches that only contain one closed figure. I gather you can create a "master sketch" and selectively project geometry into other sketches. But the last few times I've tried the app, I haven't gotten far enough into my sketches before rage quitting to validate the technique.
Right now I am back F360 with their hobby license wanting to escape their regular messing with the terms and conditions.
> Even selecting things in the UI has sucked. I went in and increased the selection radius or whatever, that helped. But really, should I need to do this as a new user?
Agree - selection isn’t broken, but it’s definitely sometimes frustrating and as it’s such a common function, absolutely should be as close to perfect as possible. I think it’s partly that the visual indication of what you’re hovering over and would be selected is too subtle, and also I’ve found (on Mac; I’ve not confirmed on other OSs) that it’s not selecting what’s at the exact tip of the pointer, but is rather selecting a couple of pixels away.
> Getting the constraints to behave is like pulling teeth.
Huh, once I’ve actually selected correctly, I find the constraints are fine - say, 95% as good as Solidworks.
> It also kind of sucks that you have to have really sparse sketches that only contain one closed figure.
Can you explain what you mean by this? Do you mean you can’t have a sketch with (to take a very simple example) a circle inside a circle, or two unrelated circles, or something else?
There are situations I can think of where selection does seem broken by design. It's fairly easy to get into a situation in the 3d view where you want to select a vertex but because of the draw order it's very hard to find an orientation of the model that lets you put it "in front". So you spend ages selecting the lines around it, spinning the model, trying again from all sorts of angles. Heaven help you if you're trying to select a bunch of points that have this problem, it's frustrating as hell. The second is in sketches, where the constraint icons aren't selectable when they're grouped but will block the selection of a component underneath them anyway. That's just obnoxious. I think in both cases the UI is working as designed, but it makes for an unusable outcome.
Oh, and if the selection point isn't at the pointer point? That's just a bug, and needs to be fixed. I can't see any defending that.
I’ve tried FreeCAD multiple times, but I’m just too used to Rhino 3D. Unfortunately, it’s rather expensive. Anyone need a slightly used, low-mileage, one owner soul?
FWIW, I've been kind of kicking myself for being bamboozled into buying a Plasticity Studio license --- really, really wish I'd instead bought Moment of Inspiration 3D (which is by the lead developer of Rhino 3D).
Curious what folks think of MOI3D (and if there is any other similar CAD tool other than maybe Shapr3D).
Recent freecad is pretty decent. My main complaint these days is the performance of the geometry engine.
I found FreeCAD fine but it takes me a while to remember how to use it each time (since I don't use it frequently...)
FreeCAD would benefit from effective
(1) agile Product Management,
(2) Product Design & continuous user-research,
(3) Improvements to test-driven development (TDD),
(4) transparent & open outcome-based roadmap,
(5) a vision to make the application easy to use for newbies in a maker-space, and (this is specific to my use-case),
(6) Improvements to the CAM module to make it easy to use this for CNC routers, and designing objects with sloped/curved surfaces.
- FreeCAD site: https://www.freecad.org/
- FreeCAD code: https://github.com/FreeCAD/FreeCAD
- FreeCAD forum: https://forum.freecad.org/
To echo others' comments: FreeCAD has improved significantly since v1.0, so I'm hoping this attracts quality & stability-minded develeopers, and a frequent release cadence.
Regarding the last point in his FAQ:
Ofc. Let's reinvent the wheel, b/c improving the existing SW would be "too much effort", and btw NIH.
Doofus.
" Hello everybody out there using minix -
What's that guy thinking... Doofus.Its not often you see 'fillets and chamfers' are tip-line features in the readme for CAD packages. But good on you for building something.
> Its not often you see 'fillets and chamfers' are tip-line features in the readme for CAD packages.
Well the readme states the following:
Solvespace on the other hand gets the workflow part right, but falls short by not importing STEP and the geometry kernel not supporting chamfers and fillets.
So I assume that's where that comes from.
This is honestly the first thing I look for with anything new claiming "CAD".
Roughly every other week there is a new "The (programmable) CAD that fixes everything!" post on the front page, just for me to open them up excitedly and noticing that they use a mesh kernel and will thus never be able to provide fillets and chamfers painlessly (for the user). All while they are absolutely essential for a lot of designs, especially in 3D-printing, a well-placed fillet/chamfer can make the difference between an object that breaks upon looking at it funny and one that can bear significant load.
Yeah, it's a tough row to hoe --- I've been handling it quite differently in my own project, taking "Design for Manufacturing" to the ultimate conclusion and requiring that the part be designed by actually applying tooling:
https://github.com/WillAdams/gcodepreview
Needs a full-on re-write to make the G-code export (and import) work well, but lately I've been focused on exporting to DXFs w/ colour/layer tagging which was just added as an import feature in the CAD tool my side-gig employer does.
G2 fillets are the next frontier. Even Fusion doesn't seem to handle them well. I mean they can be created without much drama, but geometry derived from them is very likely to fail with no real reason.
Well, implementing fillets and chamfers is no easy task, so it's well deserved to be there.
Source: been there, done that.
fillets and chamfers are at the same time both ridiculously difficult and ridiculously important.
It has been one of the main complaints about openscad for some time
Hull and Minkowski operations sort of allow one to do this sort of thing, but add to the mathematical complexity in what is probably the worst-possible way....