Full Content:
Autodesk shipped IFC 4.3 export capability in Revit 2026, responding to European markets where the updated specification is increasingly mandated for public projects. After three weeks of testing across architectural, structural, and MEP workflows, the implementation reveals Autodesk’s familiar pattern: announce compliance, ship half-baked functionality, fix it in service packs.
What Actually Works
Geometric export is solid. Wall assemblies, structural framing, and basic MEP components translate to IFC 4.3 without the coordinate system disasters that plagued early IFC 4 implementations. We tested round-trip workflows through Solibri, Tekla, and ArchiCAD—geometry survived intact in 90% of test cases.
Spatial hierarchy also improved. Building stories, zones, and space boundaries export correctly, preserving the organizational structure crucial for facility management handoffs. This is genuinely better than IFC 2×3 and addresses real practitioner pain points.
Where It Falls Apart
Property mapping is a disaster. Custom parameters—the actual data most firms care about—fail to export in 40% of test cases. We created 50 custom parameters across walls, doors, and mechanical equipment following Revit’s standard workflows. Only 30 made it through IFC export with correct values and mappings.
The problem isn’t random. Specific parameter types fail consistently:
- Yes/No parameters → Export as empty strings
- Formula-driven parameters → Export as static values (breaking downstream calculations)
- Shared parameters with custom units → Unit conversion errors in 25% of cases
MEP systems got the worst treatment. Equipment connections, system classifications, and flow parameters export inconsistently. A simple chilled water system with 12 components exported with only 7 properly classified. The remaining 5? Generic “IfcFlowSegment” with no system attribution.
The European Market Reality
Why does this matter? Because European public procurement increasingly requires IFC 4.3 for handoffs between design, construction, and facility management. Countries like Finland, Norway, and the Netherlands are mandating it contractually. For firms serving these markets, “wait for SP2” isn’t an option—they need functional IFC 4.3 now.
Autodesk knows this. They shipped minimal viable compliance to check the regulatory box, knowing European firms are captive customers who can’t easily switch platforms. It’s the same playbook they used for IFC 2×3 in 2012: ship broken, fix slowly, blame users for “improper workflows.”
What Changed from 2025
Revit 2025’s IFC 2×3 export is mature and reliable after years of refinement. The 2026 upgrade to IFC 4.3 introduces new bugs while maintaining most 2×3 capabilities. You can still export 2×3 if needed—Autodesk didn’t remove backward compatibility.
The real question: Is 4.3 worth the stability hit? For most firms, no. Unless you’re contractually required to deliver IFC 4.3, stick with 2025’s proven 2×3 export until Autodesk works out the issues.
Testing Methodology
We tested using three project templates:
- 50,000 SF office building (architecture + structure)
- Hospital MEP retrofit (mechanical, electrical, plumbing)
- Mixed-use residential (full disciplines)
Each project exported to IFC 4.3, imported into Solibri Office for validation, then round-tripped back to Revit. We measured:
- Geometry preservation (wall thickness, element placement, void handling)
- Property mapping accuracy (standard + custom parameters)
- System/assembly relationships (especially MEP)
- Performance (export time, file size)
Results were compared against identical IFC 2×3 exports from Revit 2025.
The Service Pack Gamble
Autodesk typically ships major service packs 4-6 months post-release. If history repeats, expect meaningful IFC fixes in SP2 (April/May 2026). SP1 will likely address crashes and critical geometry bugs but skip the property mapping issues.
For firms on subscription (everyone, basically), the calculus is simple: Wait unless forced to upgrade. European firms working on public contracts don’t have this luxury—they’re stuck debugging Autodesk’s incomplete implementation while under contractual deadline.
What Autodesk Should Have Done
Ship IFC 4.3 as an opt-in beta feature in 2025, gather feedback from power users, fix the obvious issues, then make it default in 2026. Instead, they rushed compliance for marketing purposes, leaving practitioners to find the bugs in production.
This pattern exhausts goodwill. Firms are increasingly evaluating alternatives (ArchiCAD, Vectorworks) not because Revit’s core functionality fails, but because Autodesk’s quality control has deteriorated while subscription prices have increased 40% since 2016.
Bottom Line
IFC 4.3 in Revit 2026 is compliance theater. It technically meets specification requirements while failing practical workflows. If you’re not contractually obligated to use 4.3, stay on 2025. If you are, budget extra time for manual data cleanup and consider parallel workflows in Solibri or Simplebim to fix Autodesk’s export failures.
The tragedy is that IFC 4.3 genuinely improves on 2×3 for infrastructure and MEP coordination. Autodesk’s implementation just isn’t ready for production use.