Evergreen

Evergreen is an experimental Common Lisp implementation written in Rust with a tiered JIT and garbage collector.

Evergreen Common Lisp is an experimental Lisp implementation written in Rust. It features a HotSpot-inspired tiered JIT, a moving generational garbage collector, and tools for delivering standalone applications. The project explores AI-driven development with minimal manual code review. It includes adaptive native compilation, optimization without required type declarations, and lightweight concurrency with fibers. Users can deliver standalone applications, call C libraries through the FFI, use Java through egcl-jvm, or access the Python CPython API. It implements macros, CLOS, conditions, and restarts, and supports ASDF systems.

Evergreen at a glance

Best for: Software engineer
Runs on: Linux, macOS, Windows, Android, CLI, open source
Pricing: Free
Released: October 2026
Category: AI coding tool
Website: github.com/atgreen/evergreen

Key features

Adaptive native compilation with tiered JIT
Moving generational garbage collector
Optimization without required type declarations
Lightweight concurrency with fibers
Standalone application delivery and saving
Foreign function interface for C, Java, and Python

Use it for

Run interactive Common Lisp REPL sessions
Deliver standalone Lisp executables
Load and manage ASDF systems
Execute cross-platform target builds

Alternatives to Evergreen

jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
LangChain4j — Integrates LLMs and vector stores directly into Java applications.
Shimmy — Runs local AI models with zero Python or C++ dependencies.
BeautifulRust — It improves agent output using guidelines from five major Rust sources.
jscpd — A native Rust binary that runs fast without a Node.js runtime and integrates with AI agents.
zerostack — Minimal memory footprint and high CPU efficiency compared to JS-based agents.
daidocs — It keeps code graphs as plain text for lower token costs and better accuracy.
ort — Provides an ergonomic Rust API for running high-performance ONNX models.
Gtree — Runs multiple AI coding agents together and flags file overlaps.
Amber 8 — It keeps execution deterministic while using a language model to handle messy text input.
Coop — Runs AI coding agents in isolated virtual machines to protect host systems.
TS Rust — It runs type checking twice as fast as the TypeScript Go compiler.

All Evergreen alternatives