🗺️ tilegen

A fast, multi-threaded tool to slice large images into map‑style tiles at multiple zoom levels.

Features

• Automatic zoom‑level calculation: Dynamically determines how deep to go based on image dimensions and a configurable magnification factor.
• Worker‑thread pool: Fully utilizes all CPU cores via worker_threads.

Installation

NPM

# npm
npm install tilegen -g
# yarn
yarn add tilegen -g
# pnpm
pnpm add tilegen -g
# bun
bun add tilegen -g

Source

Before installing, make sure you have Bun installed on your system.

git clone https://github.com/itsbrunodev/tilegen.git
cd tilegen
bun install

Usage

Help

To get help, run the following command:

tilegen -h

CLI

tilegen can be used from the command-line with the following options:

Option	Description	Default
-t, --tile-size <TILE_SIZE>	The tile size of each image.	256
-f, --tile-format <TILE_FORMAT>	The format of the output tiles.	webp
-m, --max-mag <MAXIMUM_MAGNIFICATION>	The maximum magnification factor.	1
-i, --input <INPUT_PATH>	The path to the input image.	./input.png
-o, --output <OUTPUT_DIR>	The path to the output directory where the tiles will be saved.	./out/

Example usage

tilegen -t 512 -f webp -m 4 -i ./input.png -o ./out/

This will generate tiles at ./out/{z}/{x}/{y}.webp, with a tile size of 512 and a maximum magnification of 4 from the input image ./input.png.

What is maximum magnification?

MAXIMUM_MAGNIFICATION controls how deep your zoom levels go. It adjusts how many source‑pixels each tile covers at the deepest level.

Increasing it by 1 gives you one extra zoom‑in level (more detail at deepest zoom), at the cost of more tiles and storage.

For example, if MAXIMUM_MAGNIFICATION is 4, at the deepest zoom level, each tile will cover 4x4 = 16 source pixels. If the input image is 1024x1024, this means there will be 4 zoom levels in total (2^4 = 16).

Demo

soon

How it works

I've made a post on dev.to which you can check out to see an in-depth explanation of how this tool works.

License

tilegen is under the MIT license.