Your data, instantly visualized.

Luminon

Beta

The intuitive dashboard builder that transforms raw data into beautiful browser-based interactive dashboards

npx -y @luminondev/mcp-dashboard start
Works with Claude Gemini Codex Perplexity
Open-source project
Get started View Preview
Luminon Claude screenshot

An MCP for web dashboards.

Luminon is a Model Context Protocol server for teams that want to turn raw data into working dashboard apps through an AI assistant. You connect Luminon to MCP clients like Claude, Gemini, Codex, or Perplexity, provide a dataset, and the assistant uses Luminon tools to create web dashboards with charts, filters, layouts, and updates—while the LLM can also perform the data analysis on your behalf. It is built for developers, product teams, analysts, operators, and non-technical users who need fast internal dashboards without writing frontend code or having deep dashboard tooling expertise.

Live preview

Build Visualizations faster

Features

Zero-code dashboards.
Turn CSV files or JSON rows into production-ready dashboards with natural language, auto layout, and grid-based editing.
Smart visualizations.
Build bar, line, area, donut, grouped bar, stacked bar, scatter, radar, table, and KPI card views from the same dataset.
Themes, layouts, and presentation.
Customize themes, tune grid layouts, and switch dashboards into presentation-ready views for cleaner demos and handoffs.
Filters, recovery, and dataset updates.
Apply global and chart-level filters, restore dashboards from snapshots, and replace or append dataset data as your source files evolve.
MCP server and web renderer.
Run a Node.js + TypeScript MCP server for AI tool integration, plus a React-based renderer with Nivo charts and react-grid-layout for the dashboard UI.
API without AI tokens.
Use the REST/SSE API (Node.js, TypeScript) to update datasets and dashboards directly—register, replace, or append data, and refresh layouts—without spending AI tokens when you already know the changes you need.
Product screenshot

How it works

From install to insight in four steps

  1. 1 Start with the quick install guide, then connect Luminon to Claude, Gemini, Codex, or Perplexity.
  2. 2 Provide your dataset.
  3. 3 Let your AI assistant generate the dashboard; the LLM can also perform the data analysis for you if you want.
  4. 4 Instantly visualize your data.
Luminon dashboard preview

Subscribe for news and updates

Get the latest Luminon releases, feature drops, and tips directly in your inbox.