Skip to main content

Media, Vision & Viewport Tools

These tools let the AI generate or analyze images, synthesize spoken audio, capture screenshots of your Grasshopper canvas or Rhino viewports, and navigate your canvas view to point out specific components.

Image generation, vision, and speech​

ToolWhat it does
text2imgGenerates an image from a text prompt using an image-capable AI model (supports size, quality, style, and aspect-ratio options).
img2textDescribes or answers questions about an image (from a URL, local file, or base64 data) using a vision-capable AI model.
speech_generateSynthesizes spoken audio from text using a text-to-speech model (with optional voice and playback speed).

Capturing canvas and Rhino viewport screenshots​

ToolWhat it does
canvas_screenshotCaptures the currently visible Grasshopper canvas as a PNG image (automatically sized for AI vision analysis) and optionally saves it to disk.
viewport_screenshotCaptures the active or named Rhino 3D viewport as a PNG image and optionally saves it to disk.
canvas_hi-res_screenshotRenders a high-resolution image of the Grasshopper canvas (the whole document, current selection, specific GUIDs, or custom bounds, with configurable scale and optional transparent background) for documentation or publishing.
Screenshot privacy

Screenshots capture whatever is visible on your canvas or Rhino viewport. When using the MCP Server, screenshot tools are disabled by default unless you enable Allow External Tools.

Guiding your view on the canvas​

ToolWhat it does
canvas_pointPans and zooms your Grasshopper canvas to a component or region and briefly highlights it — allowing the assistant to visually point at the components it is discussing without modifying your definition.
canvas_viewControls the Grasshopper canvas viewport (zoomExtents, frameGuids, setZoom, setCenter). On the MCP Server, this tool requires Allow View Control to be enabled.