Need to get familiar with the AEC software market? Get our report

AEC software is getting an AI upgrade right now, from the biggest vendors to brand new startups.

We've found a lot of interesting signals. For example, in June 2026, Andreessen Horowitz led a $50M Series A in Endra, a startup that uses AI to design mechanical, electrical and plumbing systems inside building models. And at its September 2026 user conference, Autodesk showed new AI tools for Revit, like a drawing check that flags missing dimensions or walls placed too close together. Investors and incumbents are both betting that AI can take over the repetitive part of design work.

In this section, we look at which design tasks firms hand to AI first, why engineering is drawing new money, and what the incumbents' moves mean for smaller software makers.

Design firms sell hours, and too many of those hours go into redrawing, checking and coordinating models.

We compare six kinds of AEC software companies, from BIM and design tools to engineering analysis, MEP design and project collaboration. Their buyers range from architecture and engineering firms to contractors and building owners. Then we rank their biggest pains, like clashes between models, rework when the design changes, and too few engineers for the work coming in.

This shows which tasks a firm will pay to speed up, and which ones it still wants a licensed engineer to handle.

Subscriptions per seat, enterprise agreements, cloud credits used by the project, paid add-ons for analysis and training: AEC software companies charge in many ways.

We look at who pays, and what each customer costs to serve. Design tools sell seats to firms that bill by the hour, so time saved can mean fewer seats. Cloud platforms charge by project or by usage and pay for computing and storage. Engineering analysis tools can charge more because one error is costly, but they sell to a smaller group.

You'll see which models still grow when AI makes each designer faster, and which ones lose seats as a result.

A growth rate tells you little until you see what drives it. So we go through the forces one at a time.

Some are easy to see: a shortage of engineers, heavy spending on data centers, energy and infrastructure, and owners asking for digital models at handover. Others reinforce each other. More projects in BIM give AI more data to learn from. Cloud platforms make it easier for everyone to work on one model. And each firm that uses AI on one task soon finds the next task to automate.

Before building or investing, you should know what convinces a design firm to change tools it has used for twenty years, and what keeps it paying afterwards.

Autodesk and Trimble are the names most firms know in AEC software, but the market goes far beyond them.

Many of the most interesting players stay under the radar. Some test site layouts for developers in minutes. Others render designs in real time, check drawings against building codes, or design homes for builders who sell the same plan many times.

They're hard to compare because each one serves a different trade. A tool for architects, one for structural engineers and one for contractors may all count as AEC software, but they sell to different people, work with different files and get judged on different results.

Below, you'll find the market grids and ecosystem maps we've built to show who serves which trade, and where new players are finding room.

Chaos, TestFit, Higharc… a few AEC software companies have found work that firms are happy to pay for.

Chaos built its rendering tools into the design software architects already use. TestFit helps developers see what fits on a site before they hire an architect. Higharc gives homebuilders one platform to design, sell and build their homes. Each found a clear use, though none has yet shown it can grow into a full design platform.

We'll look at how these teams picked a narrow task, fit into the tools firms already had, and grew from there without taking on the incumbents head-on.