Anvaka GitHub io City Roads: visualize a city’s road network

How to get started

1. Open the official City Roads page and enter a city with its country or region.
2. Choose Find City Bounds and inspect the returned geographic areas.
3. Select a city-sized boundary and wait for its roads to load.
4. Check the extent and visible attribution before comparing the network.
5. Open Customize to inspect line, background and label colors.
6. Choose PNG or SVG for your intended use, then inspect the saved output; exports were not tested in this review.

From a city name to a selected boundary

Start by entering a place name and choosing Find City Bounds. The results represent geographic areas; similarly named or overlapping administrative boundaries can produce different extents. In our test, Paris, France returned several entries and we selected one administrative result. The loaded address recorded areaId=3600007444. The selected boundary matters as much as the city name when you compare two maps.

OpenStreetMap data, name lookup and caching

The official README explains that city-name resolution uses Nominatim, while road data comes from OpenStreetMap through Overpass when it is not already in the project’s cache. The author describes a cache of roughly 3,000 larger cities to reduce downloads. These components have different roles: name search identifies the area, the data request obtains the road geometry, and WebGL draws it in your browser. An available map is not proof of a live or complete road survey.

What we verified with the Paris example

On October 25, 2026, we searched Paris, France, chose an administrative boundary and waited for the cached data to load. The application rendered a dense road network, displayed Paris as the label and retained OpenStreetMap attribution. Opening Customize revealed line, background and label color controls plus PNG and SVG export links. The screenshots record that actual session. We did not verify every road, compare boundary variants or complete a file-export test.

Custom appearance and raster versus vector output

The interface provides separate color controls for roads, background and labels. PNG is offered as a raster export of the current screen; SVG is offered as a vector export. Choose a view and contrast level before exporting rather than assuming a file will include an entire city at unlimited detail. Inspect the resulting file in the program where you will use it. Export controls and their implementation were inspected, but file dimensions, print quality and downstream editor compatibility were not tested.

Loading limits and large geographic areas

The project warns that rendering is constrained by browser and graphics-memory capacity. Its interface notes that large cities can require more than 200 MB of transfer, and the README warns that areas with millions of roads can crash a page. Overpass requests can also be slow or rate-limited. Begin with a city-sized boundary, allow the request to finish and avoid treating a stalled download as proof that the city has no roads.

Open source, attribution and developer extensions

The repository publishes the application code under the MIT license and documents a scene API for developer use. The road data has separate OpenStreetMap attribution and licensing requirements. Preserve the displayed data credit when preparing a shared visual, and consult the linked source information before adapting data or code for another project. A browser rendering, a software license and a dataset license are distinct parts of the workflow.