Lithe — a lightweight Java IDE v0.1.14 · 2026.08.11 macOS 13+ · Windows

Write Javawithout payingfor the IDE in RAM.

Lithe keeps the parts of IntelliJ IDEA you actually use every day — project tree, search, run, debug, diff — in about 1/5 of the memory.

Download for macOS Download for Windows GitHub ↗

Java & Spring Boot · Maven and Gradle

Press twice to search this site, the way you would in the IDE.

order-service
OrderApplication
src/main/java›com.example.order›OrderService.javaUTF-84 spacesJava 21

Fig. 01 The workbench, live. Open a file, search the project, or try the tool bar on the left.

A Java IDE you can leave open all day.

Browsing code sits at 200–300 MB. A normal working session runs around ≈400 MB, with peaks of 700–800 MB when services are running.

  1. Light enough to forget

    Around ≈400 MB in everyday use, so Lithe can stay in the background instead of being closed to free up RAM.

  2. The IDEA core you know

    Project structure, editor, search, navigation, and diff. Maven and Gradle projects open as they are.

  3. Ready for AI-written code

    When an AI tool edits your files, read the diff in the file it belongs to before you commit.

One fifth, measured.

Same Spring service project, resident usage after indexing completes. Every square is 16 MB; the whole field is what IntelliJ IDEA holds.

1 square = 16 MB IntelliJ IDEA · 2 GB
Lithe Range of the reading IntelliJ IDEA

Your numbers will move with project size and how many services you keep running.

How we build Lithe.

Two constraints guide every feature before it reaches the workbench.

Resource discipline

Every megabyte has to earn its place.

Language services, terminals, build tools, and database helpers start only when the work asks for them.

Baseline memory
300–400 MB
Tools and services
On demand

Workflow continuity

Keep the habits. Remove the weight.

Project navigation, search, Maven, Gradle, run, debug, and side-by-side diff review stay where Java developers expect them.

Core workflow
IDEA-style
Platform builds
Mac + Win
  • James Whitaker Backend engineer

    Project tree, navigation, and diff sit where I already look. Switching over barely took any relearning, and the memory drop is obvious.

  • Priya Nair Spring Boot developer

    IDEA plus Docker used to fill the machine. I can leave Lithe open all day now, and the fans barely spin.

  • Daniel Reeves Software engineer

    The editor and file tree feel familiar. For day-to-day Spring work I do not need a full IDEA instance sitting in the background.

  • Aimee Park Independent developer

    Search and jump-to-definition are where I expect them. Walking an unfamiliar service is much easier when the IDE stays out of the way.

  • Marcus Grant Systems engineer

    After an AI agent touches several files, the side-by-side diff stays next to the code. I can check the result before committing.

  • Amir Shah Full-stack engineer

    Everyday use sits around 400 MB. Browser, containers, and a local model can finally stay open together.

  • Hannah Cole Platform engineer

    This is the Java IDE I am fine leaving open all day. It is light, but the daily workflow is still there.

  • Marta Silva Java developer

    Existing Maven projects open as they are. I spend less time configuring the editor and more time reading the service.

It began as a solo.
Now it’s an ensemble of 33.

Every pull request merged into Lithe punched one hole in this roll. Each person plays their own pitch — the later they joined, the higher it sits — and time runs left to right. Press play to hear how it grew.

Play the ensemble 0:00 / 0:48

08.06 0 / 611 merges

AugustSeptemberOctober Today · 10.09 08.1008.1708.2408.3109.0709.1409.2109.2810.05

Swipe to follow the roll · Hover a hole to see its pull request, click to open it. Click the roll anywhere else to play from there.Tap a hole to see its pull request, tap again to open it. Tap the roll anywhere else to play from there.

FAQ

A few things worth knowing.

Short answers about Lithe, compatibility, and getting started.

01What is Lithe?

Lithe is a lightweight Java IDE for developers who keep an IDE open all day but do not need a full heavyweight environment. It keeps the IDEA-style workflows you use most — project navigation, editing, search, Maven, run and debug, Git diff — and starts everything else only when you ask for it.

02How much memory does it actually use?

Around 200–300 MB once a typical project is open, and roughly ≈400 MB in normal day-to-day use. Running several services at once pushes the peak to 700–800 MB. Actual usage varies with project size and which services you have running.

03Is it a replacement for IntelliJ IDEA?

For everyday reading, editing, navigating, running, and reviewing Java code, it is meant to be. For deep refactoring, the full inspection set, and the plugin ecosystem, IDEA is still ahead — that is a deliberate trade, and it is what keeps Lithe small. Many developers keep both and reach for IDEA when they need the heavy tooling.

04What does it support?

Java and Spring Boot are the focus, with Maven and Gradle projects opening as they are. Releases bundle the Eclipse JDT Language Server, so there is no separate setup. Entry points are auto-detected for Spring Boot, Java, Maven, Gradle, npm, Cargo, Go, Python, Make, Docker Compose, and shell projects. Java features need JDK 17 or later.

05Which platforms are supported?

macOS 13 or later on Apple silicon and Intel, plus Windows x64. On macOS you can also install and update through Homebrew.

06Is Lithe open source?

Yes. Lithe is developed in the open under the Apache-2.0 license. You can inspect the source, report issues, and contribute on GitHub.

07Where can I report a problem or request a feature?

Open an issue in the Lithe GitHub repository with the steps to reproduce, your OS version, and any useful screenshots or logs.

Travel light.

Free and open source under Apache-2.0. For macOS 13+ and Windows.

Download for free All releases

Or press twice to see what else this page can do.

Lithe