VitruAI + Revit
VitruAI ships as a signed Revit add-in that runs inside the customer’s Revit environment and drives the model directly through the official Revit API — geometry, parameters, schedules, sheet sets, the full document tree, read and write. It supports Revit 2023, 2024, 2025, and 2026 on Windows 10/11 and is currently live with our launch customer (a Dubai villa-compliance practice), where the typical install runs in ~30 minutes.
- Native Revit access — no exports, no IFC round-trips, no PDF intermediaries.
- Every live VitruAI agent connects through this Revit integration for direct model read/write.
- Annotations, flags, and QA comments write back into the Revit file the team already opens.
Install, requirements, and what runs through the integration.
-
Install in ~30 minutes
Drop the signed `.addin` and DLL into the Revit add-ins directory (per-user AppData or a shared network location), restart Revit, and pair the add-in with the workspace token your admin issued. The add-in adds a VitruAI ribbon panel with access to agents such as the Code Compliance Agent and the Studio QA/QC Agent. Typical firms complete pilot installs on 3–5 machines in a single afternoon, without needing a full IT rollout project.
-
Revit 2023–2026 on Windows 10/11
The integration supports Revit 2023, 2024, 2025, and 2026 on Windows 10/11, using the .NET runtime bundled with each Revit version, so there is no separate framework installer. The add-in is Authenticode-signed by ADAIA Ventures Ltd, which satisfies standard Revit add-in trust prompts and group policy checks. macOS Revit via Parallels or similar is not supported; firms standardise on Windows workstations for this deployment.
-
Outbound HTTPS only — no inbound ports
The add-in initiates outbound HTTPS calls only, to a single fully qualified domain name with a documented IP range, so firewall rules stay simple and audit-friendly. No inbound ports or VPN tunnels are required; traffic patterns mirror other Revit-connected tools such as Dynamo or pyRevit. Tenant regions are provisioned in eu-central or me-central, and IT teams can pin a specific add-in version that matches their security review.
-
Per-seat or per-project licensing
Licensing activates via workspace tokens that map to either individual users or project codes, so firms can choose per-seat or per-project commercial models. Tokens are independent of machine IDs, which avoids reactivation after hardware refreshes or VDI pool changes. Signed updates auto-check on Revit launch, and if a new build conflicts with a critical plugin or template, teams can roll back with one click from the VitruAI ribbon while IT investigates.
-
All live VitruAI agents run through this integration
The Revit add-in is the single integration layer for all live VitruAI agents, including the Code Compliance Agent, the Studio QA/QC Agent, BOQ/Take-off, Pre-sales Drawings, and Project Memory. Each agent reads and writes the active document via the official Revit API, so they respect worksharing, element ownership, and design options. Workflows such as Dubai Villa Code Compliance run inside the same Revit session the team already uses for modeling and documentation.
Revit integration — common questions
-
Does this work on cloud-hosted Revit (BIM 360 / ACC)?
Yes, the add-in works with cloud-hosted models on BIM 360 and Autodesk Construction Cloud because it authenticates through the same channels Revit itself uses. Cloud workshared models behave the same as local files from the agent’s perspective, including element borrowing and sync-with-central. Firms already running Dynamo or pyRevit against BIM 360 projects see similar network and permission behaviour with this integration.
-
Do we need to whitelist VitruAI in our IT environment?
Most firms add a single outbound HTTPS rule to allow the VitruAI tenant FQDN and documented IP range, with no inbound ports or VPN required. This mirrors how they already treat trusted SaaS tools that connect from Revit, keeping firewall configuration straightforward for security teams. If your firm runs strict egress controls, we provide a short whitelisting memo and sample firewall entries for review.
-
What happens if Revit crashes mid-run?
Every agent write runs inside a Revit transaction, so if Revit crashes or the user kills the process mid-run, the transaction rolls back atomically. No partial element changes or half-written parameters remain in the model, and worksharing central files stay consistent. On restart, the add-in offers to rerun the last task or open the log so the team can confirm which views, sheets, or elements were already processed.
-
Can we self-host?
Yes, self-hosted deployments are available as a Labs engagement for firms that need data residency or regulatory control beyond the standard eu-central or me-central tenants. In those projects, the same signed Revit add-in points to an on-premises or private-cloud endpoint that your IT team manages. This pattern suits larger organisations that already self-host tools like internal Dynamo packages or pyRevit extensions alongside compliance workloads.
-
How are updates pushed?
The `.addin` checks for new signed builds on Revit launch and prompts users when an approved version is available for their workspace. Updates install in place without a separate MSI, and change logs highlight any Revit API changes that might affect templates or companion tools. If a new build conflicts with a critical deadline, IT or project leads can pin the previous version so teams finish the submission before adopting the update.