Platform / Features

Two kinds of AI in ShapeDiver

AI shows up in ShapeDiver in two distinct ways, and they serve different people. One is AI inside the apps you build: an assistant your users can talk to. The other is AI that helps you build: coding agents that write your ShapeDiver front end. This page covers both, starting with the one people ask about most.

The in-app AI agent is in preview.

A ShapeDiver configurator of a wooden slatted lounger, with the AI assistant's chat panel open beside the 3D model, listing what it can do: explain the settings, change options, adjust the camera, and check model metrics.

The two kinds

AI in your apps, and AI that helps you build

It’s worth separating them, because they’re for different jobs. AI in your apps is for the people who use what you publish: an in-app agent they can chat with to drive a model. AI that helps you build is for you and your developers: coding agents that write ShapeDiver front ends for you. The rest of this page takes each in turn.

Diagram: two kinds of AI - in your apps, a user asks the in-app agent for a change and the model updates; when building, you ask your coding agent and it writes the front end

AI inside your appsPreview

An assistant your users can talk to

ShapeDiver’s in-app AI agent lets the people using your app change a model by describing what they want, instead of hunting through controls. Through a chat interface, the agent can adjust parameters, move the camera, and more - turning a plain-language request into the same actions someone would otherwise take by hand. It’s a new way in for people who know what they want but not which slider to reach for.

Note

The in-app AI agent is in preview - newly available, with more to come.
Concept illustration: a shopper types “Make it a bit longer and show me the walnut option”; the agent sets the length, sets the material, and frames the table, which updates in the viewer
Concept illustration: a shopper configuring a table could type “make it a bit longer and show me the walnut option,” and the agent sets the parameters and frames the result.

You decide what the agent can do

The agent’s tools are defined in Grasshopper

You decide what the agent is allowed to do. Its tools are defined in Grasshopper, the App Builder way, so the actions it can take are exactly the ones you wired up: which parameters it can change, what it can do in the viewer, and where it should stay out. And what it manipulates is your model, rendered in the ShapeDiver Viewer, so the results are the real thing, not an approximation.

See App BuilderThe 3D Viewer the agent drives
Diagram: the tools the in-app agent can call are Grasshopper components you define (set length, set material, move camera); anything you do not wire up, such as pricing logic, stays out of its reach

AI that helps you build

Your coding agent, made ShapeDiver-fluent

On the building side, your own coding agent (GitHub Copilot, Claude Code, Cursor) can build a ShapeDiver front end for you. ShapeDiver’s open-source Agent Skills give it expert knowledge of the Viewer API, so the code it writes uses the API correctly. The Agent-assisted build page covers this in detail.

See agent-assisted build

Where AI connects across ShapeDiver

The pieces this builds on

App Builder

Where you wire the in-app agent’s tools, no code.

App Builder

3D Viewer

What the agent manipulates, and the API a custom front end drives.

3D Viewer

AI

Put AI to work in a real model

Start a free trial and try the in-app agent on one of your own models - or talk to us about where AI fits in what you’re building. The docs cover the setup as the agent’s preview expands.

Talk to our team

Your Business

How can we help?