VitruAI + pyRevit
The VitruAI for pyRevit integration ships as a pyRevit extension that adds agent-call functions to the pyRevit Python environment plus a ribbon UI, on pyRevit 4.8+ for Revit 2023, 2024, 2025, and 2026. It is currently shipping with 1–3 design partners running Python-led automation pipelines where pyRevit scripts and ribbon buttons now call VitruAI agents directly.
- Call VitruAI agents from any pyRevit script with `from vitruai import agents`, passing Revit document context and parameters in a few lines of Python.
- Use the VitruAI ribbon that ships as a pyRevit extension so non-scripting teams can trigger common agent calls from buttons tied to standard Revit workflows.
- Keep existing pyRevit extensions and menus intact while adding agents as a Python library, so the firm’s current automation stack stays in place without refactoring.
Install, requirements, and what runs through the integration.
-
Install via pyRevit’s extension installer
Point pyRevit’s extension manager at the VitruAI extension URL, sync, and restart Revit; the VitruAI for pyRevit ribbon and Python package then load with your other tools. During first run you paste in the workspace token your admin issued, which binds the extension to your firm’s VitruAI tenant used across VitruAI + Revit and other integrations. BIM managers can deploy the extension by profile so pilot teams trial agents before wider rollout.
-
pyRevit 4.8+ on Revit 2023–2026
The integration supports pyRevit 4.8 or later running on Revit 2023, 2024, 2025, and 2026, matching current desktop deployments at most firms. Earlier pyRevit builds that still depend on legacy IronPython 2.7 runtimes are not supported because they lack the async HTTP and packaging stack the VitruAI library expects. Firms standardising on Revit 2023+ can keep a single pyRevit baseline across production and test machines.
-
Python library plus ribbon UI in one extension
The same pyRevit extension ships both the `vitruai` Python package and a VitruAI ribbon tab, so scripting teams and production users share one install. Python-fluent staff call `agents.code_compliance()` or `agents.studio_qaqc()` directly from scripts, while project teams click buttons wired to the Code Compliance Agent or the Studio QA/QC Agent. This keeps agent logic centralised while exposing it in the right surface for each role.
-
Compatible with existing pyRevit extensions
VitruAI for pyRevit loads as another extension alongside your firm’s existing toolbars, without replacing menus or changing your current extension folder layout. Scripts in your own extensions import `vitruai` like any other Python module, so standards scripts, batch exporters, or sheet tools can call agents inline. This lets you add agent checks to existing Revit standards enforcement tools instead of rewriting them from scratch.
-
Outbound HTTPS-only agent API
The extension calls a single VitruAI HTTPS endpoint over a documented FQDN and IP range that network teams can pin in firewall rules. No inbound ports open on user machines; Revit and pyRevit only initiate outbound API calls when a script or ribbon command runs. The same HTTPS API is used by Dynamo graphs, C# add-ins, and other tools, so security review happens once and then applies across VitruAI + Dynamo and other integrations.
-
What runs through the pyRevit integration
Typical calls from pyRevit include model-wide checks, view-specific diagnostics, and element-level queries passed as JSON payloads to VitruAI agents. A standards script can send all Generic Models on workshared projects to the Studio QA/QC Agent each night, while a permit set script calls the Code Compliance Agent on selected views. Each Beta deployment ships a per-project accuracy report calibrated to the customer’s pipeline so BIM leads see how agents behave on their real models.
pyRevit integration — common questions
-
Does this require the firm’s BIM team to know Python?
The VitruAI for pyRevit integration works in two layers: a ribbon UI for non-scripters and a Python library for teams that already write pyRevit scripts. Project staff can click ribbon buttons to run checks without writing a single line of code, while computational teams embed `from vitruai import agents` in their automation. Firms often start with ribbon-driven Revit standards enforcement and then move agent calls into their own scripts as they see the value.
-
How does it differ from the Dynamo integration?
VitruAI for pyRevit runs inside the pyRevit Python environment, while the VitruAI + Dynamo integration exposes the same agents as nodes in visual graphs. Both surfaces call the same HTTPS agent API and can reach the same agents, including the Code Compliance Agent and the Studio QA/QC Agent. Firms with mixed teams often keep visual workflows in Dynamo for designers and Python workflows in pyRevit for BIM automation specialists.
-
Does it conflict with our existing pyRevit extensions?
No, VitruAI for pyRevit installs as another extension alongside your existing toolbars and scripts. It does not rename tabs, move buttons, or alter your current extension folder structure, so your custom tools keep working as before. Existing scripts can gradually import `vitruai` to add checks or automation, for example inserting an agent call at the end of a standards script that already enforces view templates and naming rules.
-
Can we call agents from a Revit C# add-in?
Yes, the underlying VitruAI agent API is HTTPS-based, so any Revit C# add-in that can make authenticated web requests can call the same endpoints the pyRevit extension uses. The pyRevit integration is a convenience layer that packages authentication, context extraction, and error handling for Python users. Firms with mixed stacks often keep C# add-ins for core UI extensions while using pyRevit scripts and Dynamo graphs to orchestrate higher-level agent workflows across VitruAI + Revit and other tools.
-
What does the Beta cohort look like?
The Beta cohort for VitruAI for pyRevit is 1–3 design partners with active Python-led automation pipelines and a maintained pyRevit extension library. These firms already run nightly scripts for tasks like Revit standards enforcement and now add agent calls into those same scripts. Beta partners work directly with the VitruAI team to prioritise which agents appear first in the ribbon and which Python helper functions land in the `vitruai` package for their workflows.