The answer to your question is in the public releases. MS went from primary partner (under ROFR) to one of the options. They retain IP rights and API hosting, although in recent weeks we learned that OpenAI was planning a workaround with AWS and Microsoft said they might sue them for that. So the happy marriage is over, it’s more like a custody battle now: https://www.reuters.com/technology/microsoft-weighs-legal-ac...
It also helps that laser beams diverge. By the time it gets back to Earth, the diameter of the beam from Artemis is probably several hundred meters, if not several kilometres. Their aim still needs to be fairly precise, but they're not trying to hit a lens with a beam that's still the width of a pencil. They really just need to paint the neighbourhood that NASA's sensors are located in.
Anthropic doesn't have anything else other than the Claude models.
But notice that no-one, not a single mention of Deepseek tells me that they are preparing to scare everyone again. Which is why Dario continues to scare-monger on local models.
Sometimes you do not need hundreds of billions of dollars for inference when it can be done locally with efficient software; and Google proved that. But where is the money in that? So continues the flawed belief in infinitely buying GPUs to scale which Nvidia needs you to do.
Only a matter of time for local models to reach Opus level. We are 1 or at most 2 years behind that and Anthropic knows that.
You have a GPU already (at least an iGPU and an NPU on most newer platforms) as part of your computer, might as well get some use out of it with local inference. And trying to do inference on a larger model with an undersized GPU will have you idling a lot less than 99% - but that still makes a lot of sense for most casual users who will only rarely need a genuine "Pro" class answer from AI. Doing that locally is way less hassle than paying for a subscription or messing with API spend.
Dune3D is more like Solvespace with a few improvements and bug fixes vs being anywhere near FreeCAD in terms of capability. Improvements include using STEP files in assemblies and having some ability to make Fillets or Chamfers. Bugs fixes would be due to using OCCT for NURBS surfaces - solvespace frequently fails with NURBS boolean operations.
As for overall capability, FreeCAD does everything these others do but also supports lofting and other modeling options, BIM for architecture, I think it does pre- and post- processing for FEA, and maybe some other "big tool" things.
Why would someone make a context switch HIGH latency? That defeats the purpose.
reply