RhinoAI has just been released for Rhino with lots of exciting and powerful new features.
It is included by default in Rhino BETA and for Rhino 8 it can be downloaded via the Package Manager from within Rhino 8.
What is it?
RhinoAI is a plugin that connects Rhino and Grasshopper (1 & 2) to AI assistants like Claude, GitHub Copilot, Codex, Gemini, and Local LLMs. You can use this to automate tasks, query for information and improve your developer workflow.
The connection works both inside, and outside of Rhino.
Features
Along with many many bug fixes, there’s lots of new and exciting features.
GH1 & GH2 Tooling
GH1 & GH2 tooling for creating, reviewing and tidying scripts. You can even use this to convert GH1 scripts to GH2.
Python & C# Scripting
Leverage the Rhino SDK’s and AI’s knowledge of coding to run scripts.
Command Creation
Allow AI to generate new commands, saved in your Rhino session, complete with icons and a shareable plugin.
AI Scripting
Create macros that run agents interactively with a prompt
Parallel Rhino Agents
The Rhino MCP Server can now run multiple Rhino agents in Parallel, this includes Rhino 8 and Rhino BETA.
AI Panel
Chat panel built into Rhino that can assist you with all of your tasks. Currently supports Claude and Codex. Run command AIPanel to toggle it on.
More Connections
A new Claude Desktop connector is available to connect Rhino to Claude Desktop.
Development capabilities
The Rhino MCP Server does not connect directly to Rhino, but instead connects via a Router. This means your AI Agent can start Rhino, load Rhino or Grasshopper plugins, test commands and components, handle and receive crash info and iterate your tooling for you. AI can debug your plugin for you from your IDE.
Example Workflows
Modeling architecture with AI → here
Modeling within an interior photograph → here
Modeling with feature recognition, reverse engineering and defeaturing a model → here
Creating and managing BIM schedule in Rhino → here
Reducing STEP file size → here
More Info
Rhino and AI → here.
Documentation can now be found → here.
Source code can be found → here.