How to Boost Your Go Application with the Green Tea Garbage Collector

Introduction

If you're a Go developer looking to squeeze more performance out of your applications, the new experimental Green Tea garbage collector (GC) introduced in Go 1.25 is a game-changer. Many workloads experience 10% to 40% less time in the garbage collector, with no changes to your code. This guide walks you through enabling Green Tea, testing it on your own projects, and providing feedback to help shape its future as the default GC in Go 1.26.

How to Boost Your Go Application with the Green Tea Garbage Collector — Source: blog.golang.org

Green Tea is production-ready and already in use at Google. Follow the steps below to start benefiting from it today.

What You Need

Go version 1.25 or later – download from go.dev/dl
A Go application that makes frequent heap allocations (most real-world apps qualify)
A build environment (terminal, IDE, or CI pipeline) where you can set environment variables
Benchmarking tools (optional but recommended to measure improvement)

Step-by-Step Guide

Step 1: Update Your Go Installation

Green Tea is available starting with Go 1.25. If you haven't already, upgrade your Go toolchain:

Visit the official downloads page and install the latest release.
Verify the version: go version should show go1.25 or higher.

Step 2: Build Your Application with the Green Tea GC

Green Tea is not enabled by default; you must set the GOEXPERIMENT environment variable to greenteagc when building. There are two ways to do this:

Option A: Command-line flag (temporary)

GOEXPERIMENT=greenteagc go build -o myapp ./...

Option B: Persistent environment variable (for repeated builds)

export GOEXPERIMENT=greenteagc
go build -o myapp ./...

Important: This flag affects only the build; the resulting binary will use Green Tea at runtime. If you omit the flag, the binary uses the default GC.

Step 3: Benchmark Your Application's Performance

To see the actual impact, compare your application's garbage-collection behavior with and without Green Tea. We recommend:

Create two builds: one with GOEXPERIMENT=greenteagc and one without.
Run both under the same workload (e.g., your standard load test or benchmark suite).
Measure GC pause times, total GC CPU time, and memory usage.

Tools like pprof and trace are helpful. For quick checks, enable GC logging with GODEBUG=gctrace=1 and compare the output.

Step 4: Analyze the Results

Look for improvements in these key metrics:

GC CPU time – many workloads see a 10–40% reduction
Total application throughput – less time in GC means more time for your code
GC pause latencies – Green Tea may reduce pauses and their frequency

Be aware that some workloads – especially those with very little heap allocation – may not benefit significantly or even see a slight regression.

Step 5: Monitor for Issues

Even though Green Tea is production-tested at Google, your application's behavior may differ. Watch for:

Increased memory usage (though this is rare)
Unexpected crashes or panics
Changes in program semantics (should not happen, but always double-check)

If you encounter any problems, file a new issue on the Go issue tracker with the title prefix greenteagc:.

Step 6: Share Your Feedback

Your feedback is vital for the Go team. Here’s how to report:

Problems – file a new issue and describe your workload, the issue, and relevant logs.
Successes – reply to the existing Green Tea issue (the exact link is in the Go blog post) with your performance results and any observations.

Tips and Best Practices

Start with a non-production environment – test Green Tea in staging or with a limited workload before rolling out to production.
Automate benchmarking – integrate the two builds into your CI pipeline to catch regressions quickly.
Combine with other Go profiling tools – use pprof to identify allocation hot spots; Green Tea may offset them further.
Stay informed – the Go team plans to make Green Tea the default in Go 1.26, so updates may change its behavior. Watch the Go Blog.
Report even small successes – the team wants to know which workloads benefit most to improve the collector for everyone.

By following these steps, you'll be among the first to harness the power of the Green Tea garbage collector and help shape the future of Go memory management.