Unciv: a Civ V remake in Kotlin that runs on a potato

What Unciv actually is, and who the README says it is for

Unciv is an open source remake of Civilization V for Android and desktop, written in Kotlin and built on LibGDX. The repository describes it as moddability-focused, and the licence is MPL-2.0. That combination sets the project's boundaries early: this is a reimplementation of game mechanics, not a re-release of Firaxis assets.

The README answers the obvious question directly. If you want high-resolution graphics, a full soundtrack and animations, it points you at Civilization V instead. The stated target is a small, fast, moddable, FOSS 4X that can still run on a potato. That sentence is the whole product decision. Everything else in the project follows from it, including the low system requirements, the aggressive release cadence (three releases in the two weeks before 2026-09-17) and the fact that the roadmap is ordered as polish first, then Gods and Kings mechanics, then Brave New World mechanics such as trade routes and a world congress.

Who is it for, then? People who already know how Civ V works and want to play something close to it on hardware that would not run the original, or who want to change the rules themselves. It is not aimed at players who want a new 4X design. The README is explicit that features absent from Civ V will not be added to the base game, possibly only as modding hooks, because a clear vision is what gets things finished.

Kotlin, LibGDX and a shared core across Android and desktop

The repository layout tells you most of the architecture. There is a core/ module, an android/ module, a desktop/ module and a server/ module, tied together by settings.gradle.kts and a buildSrc/ directory for build logic. That is the standard LibGDX split: game logic and assets live in core, and each platform module supplies the launcher and packaging. A single Kotlin codebase therefore produces the Android app and the desktop builds.

The Dockerfile makes the desktop build path concrete. It starts from accetto/ubuntu-vnc-xfce-opengl-g3, installs OpenJDK 17 and 11, downloads packr-all-4.0.0.jar and a Temurin 11 JRE tarball, then runs the Gradle tasks desktop:classes, desktop:dist, desktop:zipLinuxFilesForJar and desktop:packrLinux64 to produce a bundled Linux build. The run stage copies the result into /usr and launches /home/headless/Desktop/Unciv.sh. In other words, the desktop distribution is a jar plus a bundled JRE, which is why the README can offer a plain java -jar Unciv.jar path as a fallback.

The docker-compose.yml exposes ports 6901 and 5901 and runs the same Unciv.sh, which lines up with the VNC base image: the container gives you a virtual desktop with the game inside it rather than a headless server. There is also a server/ module in the tree, but the README does not document what it serves, so treat any assumptions about it as unverified.

Installing Unciv on PC, Android, Mac and Raspberry Pi

The README lists a package manager or store for nearly every platform, which is the practical reason to prefer it over a standalone download. Android comes from Google Play or F-Droid. Linux comes from itch.io, Flatpak via Flathub, or the AUR package unciv-bin. Windows offers an MSI from the latest release, plus itch.io, Chocolatey and Scoop. Raspberry Pi is covered by Pi-apps.

macOS goes through Homebrew. The README gives the exact two commands:

brew update && brew install unciv

After that, unciv should be on your PATH as a normal Homebrew formula. The README also links a separate install guide for macOS, which suggests Homebrew is not the only supported route.

The universal fallback is the jar from the Releases page, run directly:

java -jar Unciv.jar

The README flags this as not recommended, on the grounds that the project updates frequently and a manually downloaded jar goes out of date quickly. That is a fair warning rather than a limitation of the jar itself. If you want to build it yourself, the repository points to a Building Locally page in the docs site with a without-Android-Studio path, and the Dockerfile shows the Gradle tasks involved. For a first real session, the README says the tutorial content lives in-game under menu > civilopedia > tutorials, so there is no separate onboarding document to read first.

Mods are the point, but the README leaves the loading path to the docs

The repository calls itself moddability-focused, and the contributing section has a dedicated entry point for modders on the project's documentation site. What the README does not contain is a mod installation procedure. There is no directory name, no in-game menu path and no file format described in the text available here. Anyone arriving with the search phrase how to install unciv mods will not find the answer in the README.

That is a real gap, and it matters because mods are the project's stated differentiator. The likely reason is that mod handling is documented on the docs site and inside the game rather than duplicated in the repository root. The practical consequence for a new user is that the first step is the in-game mod interface or the modders documentation, not the README. If you are evaluating Unciv specifically as a modding platform, read the modders page before you install anything, because the base game's feature set is deliberately frozen to Civ V.

Where Unciv is the wrong tool

The README is unusually candid about the failure cases, and they are worth taking at face value. There is no iOS version and none is planned. The stated reasons are the cost of paying Apple, an extra release path, and the maintainer not owning an iOS device to test on. If your team or household is iOS-only, Unciv is not a candidate.

There is no Steam release either. The README says Steam decided not to host it, and speculates about legal caution regarding Firaxis. So if your expectation is a store-managed library entry with cloud saves, that expectation will not be met.

Feature requests outside Civ V are declined by policy. The README states that if a feature is in the original Civ V it will be implemented, and if it is not, it will not be added to the base game, possibly only as a modding capability. It also rules out Civ VI mechanics outright. And the graphics ceiling is self-declared: the project recommends Civ V itself if high-resolution art and animation are what you want. Unciv competes on size, speed and moddability, not presentation.

Unciv versus FreeCiv and the commercial 4X options

The closest open source comparison is FreeCiv, and the difference is the target, not the licence. FreeCiv is a long-running reimplementation of the Civilization series as a whole, with its own rulesets and clients, and it is not tied to one entry in the series. Unciv reproduces Civ V specifically, down to the roadmap order of Gods and Kings and then Brave New World mechanics. If you want ruleset flexibility across multiple Civilization eras, FreeCiv is the broader tool. If you want Civ V's particular hex-and-one-unit-per-tile design in an open source package, Unciv is the narrower and more faithful one.

Against the commercial games, the README makes the comparison itself: for graphics, sound and animation, play Civilization V. Against Civilization VI, the README declines to implement those mechanics, so Unciv is not a substitute for the newer design. The honest framing is that Unciv trades production values for portability and mod access. It runs where the commercial games do not, and you can change its rules without a modding SDK.

Maintenance, release cadence and what MPL-2.0 means here

The repository is not archived, and the last push was on 2026-09-17, the same day as the 4.22.1 release. The two preceding releases, 4.22.0 and 4.21.19, landed on 2026-09-14 and 2026-09-09. That is a fast cadence, and it explains the README's warning about standalone jars and APKs going out of date. If you install through a package manager or store, the upgrade cost is whatever that channel charges you; if you install from a release artifact, you own the upgrade problem yourself.

Licensing is MPL-2.0, a file-level copyleft licence. Modifications to MPL-covered files must stay under MPL when distributed, while larger works that combine MPL files with others can be licensed differently. That is a summary of the licence's structure, not legal advice; if you plan to redistribute a modified build, read the LICENSE file in the repository and get your own advice.

The README also addresses the legal question around cloning Civ V mechanics. It cites US Copyright Office FL-108 for the position that intellectual property rights do not apply to mechanics, and states that using assets from the original game (images, sound) is definitely illegal. It describes using the Civilization name or impersonating the games as probably illegal, with no solid sources given. So the code is MPL-2.0, but the asset and naming rules are a separate question the project itself leaves partly open.

Editorial conclusion

Adopt Unciv if you want an in-depth 4X that runs on low-end hardware, ships through package managers on most platforms, and treats modding as a first-class feature. Skip it if you want Civ V's production values, an iOS build, or a Steam library entry; the README says iOS is not planned, Steam declined to host it, and Civ VI mechanics are not coming. Before committing, verify two things yourself: that your platform's package really tracks the release cadence (the README warns that standalone jars and APKs go stale quickly), and that the mods you care about exist for the current version, since mods are loaded from a directory rather than a store.

Frequently asked questions

Official sources

1. Issues
2. License: MPL-2.0
3. README
4. Releases
5. yairm210/Unciv on GitHub