Run AI workflows from inside Rhino. Render screenshots, generate 3D models, and have the chat agent write Python that runs in your active document. The same install also enables Runchat in Grasshopper.

​

Getting Started:

​

Install Runchat for Rhino

• In rhino, run PackageManager
• Search for Runchat and Install it
• Restart Rhino

​

Launch Runchat and Sign In

Run the Runchat command in Rhino to open a web browser that is connected to Rhino. Log in to your Runchat account and then click on a workflow from your dashboard or create a new one to open the editor

​

Rhino Tools

​

Use Rhino screenshots in your workflows

You can interact with your Rhino document using two buttons in the Action Bar on the left hand side of the Runchat Editor. Click the Screenshot button to capture a screenshot of your current active Rhino view and load it as a node in Runchat. For the full rendering workflow, see Render from a Rhino screenshot. For a fast tour of what’s possible, watch Rhino’s AI Upgrade or Architectural Visualization in Rhino.

​

Import 3D models directly to Rhino

First, generate a 3d model in Runchat. Then select the node and click the Import button in the Action Bar. This will download the model and materials and load them in your Rhino viewport at 0,0,0. For more on 3D model selection, see Generate 3D models from images. For a deeper walkthrough on multi-view 3D, watch Image to 3D Building with Hunyuan 3D.

​

Send curves between Rhino and Runchat as SVG

Select curves in Rhino and click Import objects as SVG in the Action Bar. The geometry arrives as a vector node in Runchat that you can iterate on with a Code node or use as input in a workflow. To send back the other way, select an SVG node in Runchat and click Export in the Action Bar. The curves land in Rhino with their colours preserved. This is great for bringing AI-generated floor plan variations into Rhino as editable linework.

​

Generate and run Rhino Python scripts

To generate python scripts, just describe what you want in the chat window. For simple scripts, using the default chat model will be fine. For more advanced scripts, consider switching to a model that performs well in coding benchmarks (e.g. Anthropic models). The coding agent works best with incremental implementation - start with a small feature and iterate from there. The agent will create a script on the Runchat canvas when it’s finished. It might also make some UI for setting parameters in the script. You can press the Run button on the script node to run it in Rhino. If you need changes, continue the chat and paste screenshots from Rhino as required. For deeper Rhino scripting walkthroughs:

​

Browse saved scripts

Type RunchatScripts in the Rhino command bar to open a second panel that lists every script you’ve generated, organised by workflow. Each script keeps a version history so you can preview the code, rerun an earlier version, or copy code into a new file.

​

Using the Runchat Agent in Grasshopper (Beta)

You can also use the chat agent in the Runchat Rhino plugin window to interact with your active Grasshopper document. The agent can:

• Read the canvas and explain what the definition does
• Read the inputs and outputs of individual components
• Search for components and add them to the canvas
• Connect and disconnect components to build definitions

You can use the agent for:

• Asking questions about your grasshopper definition
• Building simple definitions from scratch
• Converting Rhino scripts to grasshopper definitions

For full setup and the API key flow, see Runchat for Grasshopper.

​

Next steps

• Render from a Rhino screenshot: generate styled renders from your existing Rhino work
• Generate Rhino scripts with AI: write Python that runs in Rhino by describing what you want
• Build parametric panelling scripts: advanced constraint-heavy script generation
• Generate 3D models from images: import AI-generated geometry into Rhino