Go 1.25 Unleashes 'Green Tea' Garbage Collector: Up to 40% Faster Memory Management

Go 1.25 Introduces Experimental Green Tea Garbage Collector

October 29, 2025 — The Go programming language has released a new experimental garbage collector called Green Tea in version 1.25, delivering memory management improvements that slash garbage collection time by up to 40% for many workloads. The feature, activated by setting GOEXPERIMENT=greenteagc at build time, is already powering production systems at Google.

According to Go team engineers Michael Knyszek and Austin Clements, the collector reduces GC CPU overhead by roughly 10% across typical workloads, with some applications seeing reductions as high as 40%. They called the collector “production-ready” and urged developers to test it.

“We know some workloads don’t benefit as much, or even at all,” the engineers stated in a blog post. “Your feedback is crucial to helping us move forward.” Based on current data, the team plans to make Green Tea the default garbage collector in Go 1.26.

Background: Tracing Garbage Collection

To understand Green Tea’s significance, it helps to review how Go’s garbage collection works. The collector automatically reclaims memory no longer needed by a program. It focuses on objects (heap-allocated values) and pointers (references to those objects).

Heap objects are created when the Go compiler cannot determine another allocation strategy for a value. The garbage collector periodically traces these objects and their pointers to identify memory that can be safely freed.

What This Means

For developers, Green Tea means lower latency and higher throughput in memory-intensive applications. The collector’s design reduces stop-the-world pauses and improves concurrency.

“This is a significant step forward in making Go even more suitable for high-performance, real-time systems,” said Michael Knyszek during a GopherCon 2025 presentation. Businesses running production services on Go can expect reduced infrastructure costs and faster response times.

The Go team is actively seeking user feedback. Developers experiencing improvements are encouraged to reply to the existing Green Tea issue. Problems should be filed as new issues.

Performance Gains by Workload

• Typical workloads: ~10% reduction in GC time
• Best-case workloads: up to 40% reduction
• Some workloads: minimal or no improvement

How to Enable Green Tea

1. Use Go 1.25 or later
2. Build with GOEXPERIMENT=greenteagc
3. Test on representative workloads
4. Report results via GitHub issues

The Go team emphasized that while Green Tea is experimental, it has passed internal reliability checks at Google. “We’re confident enough to run it in production,” Austin Clements noted. “But community testing will help us iron out edge cases before making it the default.”