Stantec leveraged Forma’s environmental analyses on cancer hospital in Dubai
September 16, 2025Upcoming 3D generative AI foundation models for Autodesk Fusion and Forma
The evolution of creative expression
In the early stages of a design project during ideation and conceptualization, tools should accommodate the broadest methods of expression. Early specifications, outlines, rough-sketches, and photos, among others, are all essential in expressing ideas. Creators need technology that does not obstruct their creative flow. Instead, tools should enable rapid and iterative exploration of ideas, regardless of how detailed or approximate those ideas might be. Over the past decade, research has underscored that such early-stage explorations significantly enhance project metrics like cost, sustainability, and suitability.
Not only have these flexible tools been lacking but attempts to develop them have resulted in solutions that cannot be transferred easily to detailed design development. What if future technology could unlock this exploration potential, generating useful CAD objects spontaneously and enabling seamless transitions from conceptual to detailed design?
The promise of neural CAD foundation models
Fast-forward to today, and we are in a transformative era in design and make technology. Picture a scenario where your computer comprehends spoken language, sketches, three-dimensional design data, and industry-specific workflows. Now, enrich this scenario with decades of your team’s project knowledge.
Autodesk is pioneering the development of a range of neural AI foundation models focused on design to make problems. The first of these is neural CAD, a category of generative AI models trained to directly reason both about CAD objects and industrial and architectural systems. Moving far beyond the practice of pairing large language models with existing solutions, Autodesk’s approach completely reimagines the traditional software engines that create CAD geometry or translate between industrial and architectural systems. Neural CAD foundation models offer a path to enabling the kinds of explorative interactions mentioned above but also promising a future of more connected design and make processes.
This week, at AU 2025, Autodesk announced the upcoming commercial availability of this new category of generative AI foundation model technology in Forma and in Fusion.
The first AEC-related foundation model developed by Autodesk, neural CAD for buildings enables customers to quickly transition between early design concepts and more detailed building layouts and systems.
This groundbreaking generative AI technology, neural CAD for geometry, can create designs spontaneously based on a text prompt. It is an entirely new, machine learning approach to generating CAD objects in contrast to the classical parametric CAD engines employed for over 40 years.
Bridging the gap: From research to reality
Autodesk founded its AI Lab in 2018 to research the creation of a unique category of AI focused on design, systems, and the physical world. Our researchers designed these neural foundation models to complement the existing large language models by introducing the ability to directly reason about CAD geometry, system designs, and the real world. The models are trained to reason at both a detailed geometry level as well as at a systems and industrial process level, exploring ideas like efficient machine tool paths or standard building floorplan layouts.
These neural CAD foundation models will also, in the future, be available to our customers to customize by tuning them to their organization’s proprietary data and processes. The integration of these models will transcend the capabilities of today’s tools, transforming the creative process and making it more intuitive, collaborative, and abundant.
The impacts and immediate future
Today’s parametric CAD systems have remained largely the same for nearly 40 years. Neural CAD is now charting a future far beyond this, opening entirely new possibilities for designing and making the world around us. The rapid evolution of this technology proves that even the most forward-looking visions from years ago can materialize swiftly. For example, consider the generative AI AutoConstrain feature in Fusion that helps designers quickly constrain their drawings, streamlining a tedious, time-consuming process. Autodesk launched that feature only seven months ago and it has already seen significant advancements. So, while neural CAD is a new concept and category, you can expect it soon and rapid improvement will follow.
Conclusion
The future that Autodesk has described for many years is here. As we stand on the brink of this transformative era, tools like language models and neural foundation models are becoming integral parts of our creative workflows. The convergence of these technologies is not just about enhancing productivity; it’s about making design and creation processes intuitive, collaborative, and accessible for all.
This future is unfolding rapidly, driven by relentless innovation and the power of integration. Let us seize these opportunities, integrating extraordinary tools into our creative journeys, and collaboratively build a future where the disciplines of design and make know no bounds.
