by NVIDIA
NVIDIA Elements MCP connects AI coding assistants such as Claude Code, Cursor and Codex to NVIDIA's Elements design system, with component APIs, design tokens and examples. It is for developers building AI infrastructure, robotics and autonomous-vehicle interfaces.
PluginContact for pricing
Step 1: Install the Elements CLI with the install script (curl -fsSL https://nvidia.github.io/elements/install.sh | bash on macOS/Linux) or npm install -g @nvidia-elements/cli.
Step 2: In your project, run nve project.setup to configure the MCP server for Cursor, Claude Code and Codex and add Elements dependencies.
Step 3: Or add it manually: an nvidia_elements entry in .mcp.json (Claude Code) or .cursor/mcp.json (Cursor) with command nve and args ["mcp"].
Step 4: Restart your assistant, or enable the server under Cursor's Tools & MCP settings.
Step 5: Ask your assistant to build UI with Elements components; it can look up APIs and validate templates as it goes.
The docs give setup steps for Claude Code, Cursor and Codex, and the server works with other MCP clients.
Yes. The NVIDIA Elements repository on GitHub is licensed under Apache 2.0.
Elements components are framework-agnostic Web Components, with integration guides for React, Next.js, Angular, Vue, Svelte, SolidJS, Preact, Lit, Nuxt, Go and plain HTML.
Install the CLI, then run nve project.setup. It detects your package manager and configures the MCP server for Cursor, Claude Code and Codex.
Requirements
FramerFramer is a website builder with a design agent that works directly on an editable canvas. Sites are published with hosting, CMS, analytics and SEO built in.
SplineSpline is a browser-based 3D design tool where teams and AI agents model, animate and publish interactive 3D together in real time. Scenes can ship to the web, iOS and Android.