What are you making a map for?

The renderer is the same everywhere. These pages differ in how you get to it, and what good looks like when you are done.

By who is asking

A story brief becomes a filed graphic. Elections, conflict, trade, breaking news.

Figures for papers and reports, reproducible when the data is revised.

One statistic, one map, pasted into the issue before it ships.

Values by country or region, binned properly, with a legend.

One GET, one SVG or PNG. No SDK, no tiles, no headless browser.

Claude, Cursor or ChatGPT making real maps through MCP.

Every example on the site is a live URL you can copy.

Coming from a tile-based static images API, and want the honest comparison?