Like Claude for Revit… but Free and Private
What if you could use AI inside #Revit without sending your #BIM data to the cloud?
In this video, I test my private AI workflow for Revit using local AI, Ollama, VS Code, Cline, #MCP, and a live Revit model. The goal is simple: can AI running on my own computer interact with Revit, understand the project, automate BIM tasks, and help create useful project outputs?
I test this workflow on a real Revit model by asking the AI to identify the open project, select and visualise doors, colour doors by level, create a door schedule, add fire rating information, export Revit data to CSV, and generate a self-contained HTML dashboard for exploring the results.
The key point is privacy. The AI runs locally, the workflow stays on my machine, and the BIM data does not need to be sent to cloud AI tools.
This is a practical test of what local AI could mean for architects, engineers, BIM managers, designers, and construction professionals working with Revit.
Nonica A.I. Connector Pro:
http://buy.stripe.com/3cscNl1Hb2iY2SQ...
Cline MCP configuration for Revit:
```json
{
"mcpServers": {
"Revit": {
"autoApprove": [
"get_category_by_keyword",
"get_parameter_value_for_element_ids"
],
"disabled": false,
"timeout": 60,
"type": "stdio",
"command": "C:\\NONICAPRO\\OtherFiles\\System\\Core\\net8.0-windows\\RevitMCPConnection.exe",
"args": []
}
}
}
```
Topics covered:
Local AI for Revit
Private AI workflow
BIM automation
Revit AI workflow
MCP for Revit
Ollama local AI
Cline and VS Code agents
AI for architecture, engineering, and construction
Door schedules, fire ratings, CSV export, and BIM dashboards
Qwen 3.6 running locally for BIM automation
Chapter:
00:00 Local AI is changing BIM
00:30 Cloud AI vs private local AI
01:17 How AI talks to Revit with MCP
02:59 The tools needed for the workflow
05:45 Why previous models failed
06:37 The model that can run locally
09:09 Controlling Revit with AI
12:36 Creating a full BIM dashboard