Bun 1.4 Officially Released: Major CPU & Memory Improvements via Rust Rewrite

AFFILIATE_PRODUCTS: (None)

The latest version of the server-side JavaScript runtime, “Bun 1.4,” was officially released on August 20, 2026. The biggest change is the complete rewrite of the codebase—which had been developed in the Zig language—from Zig to Rust. The migration from Zig to Rust, leveraging Claude Code, was completed in approximately 11 days, and this version is the first to reflect the results of that effort.

According to reporting by jniino from Publickey, Bun 1.4 fixes over 2,900 GitHub issues and adds 1,517 tests from the Node.js test suite. With the migration to Rust, the overall stability of the runtime has improved, with particularly notable enhancements in memory management behavior.

Major Improvements in Node.js Compatibility

Since its introduction, Bun has been developed as a “Node.js alternative,” but a lack of compatibility has been one of the main reasons developers hesitate to migrate to production environments. Bun 1.4 has seen the most significant compatibility improvements in Bun’s history.

Specifically, the testing frameworks Playwright and vitest now work. These rely heavily on Node.js APIs and were previously partially incompatible, making normal use difficult in earlier Bun versions. Furthermore, support for the OpenTelemetry telemetry collection infrastructure has been added, enhancing observability in production environments.

A memory leak that occurred during SSR processing in Next.js App Router has also been fixed. This issue, where garbage collection reference cycles occurred during query calls, causing memory to increase per page, appears to have been fundamentally resolved through the introduction of memory management via Rust’s ownership model.

Reduced CPU and Memory Usage

The performance improvements are clearly quantified. Idle CPU usage has been reduced to 1/5 of previous levels. Memory consumption has been cut by up to 32%. Startup speed has also been accelerated by up to 50% on Linux environments, with a reported figure of 4.8 milliseconds.

Regular expression processing speed has been accelerated by up to 281 times compared to Node.js. Regular expressions are frequently a bottleneck in JavaScript runtimes, and this magnitude of improvement is significant for workloads that rely heavily on pattern matching.

The background enabling these improvements through the migration to Rust includes Rust’s zero-cost abstractions and ownership-based memory management. While Zig also allowed for manual memory management, it is believed that Rust’s type system guaranteeing safe code at compile time has reduced runtime overhead.

Startup Speed and Overall Performance

The 4.8-millisecond startup speed is highly significant for serverless environments and CI/CD pipelines. Slow cold starts have historically been a challenge for Bun, and this improvement is a factor that could bolster adoption in Edge Computing and FaaS.

As reported in Popular site lobste.rs successfully migrates to SQLite, reducing CPU usage and halving costs, the choice of runtime and infrastructure has a major impact on CPU and memory usage, which can determine operational costs. Bun 1.4’s performance improvements similarly have the potential to directly reduce costs in Infrastructure as Code environments.

Editorial Opinion

In the short term, Bun 1.4’s improved Node.js compatibility has lowered the migration barrier for developers. The fact that Playwright and vitest now work will likely accelerate consideration of migrating existing Node.js projects. Particularly when selecting options for new projects, the clear quantification of the performance difference strengthens the case for adopting Bun.

In the long term, the development process itself—the 11-day Zig-to-Rust migration assisted by Claude Code—warrants attention. The ability to perform a large-scale, cross-language migration of a codebase with AI assistance in a short period could fundamentally change the engineering cost of runtime development. Whether similar methods proliferate in other OSS projects will likely become a key discussion point.

It is difficult to isolate and evaluate the degree to which the improvements brought by the Rust migration are attributable to the language choice versus the concurrent fix of over 2,900 bugs. Benchmark comparisons between language-specific builds, to accurately measure the effect of the Rust migration, are awaited.

References

• “Rustに移植された「Bun 1.4」正式リリース。Node.jsとの互換性向上でPlaywrightやvitestが動作、CPUとメモリの使用率が大幅改善、正規表現が高速化など”, by jniino — Publickey, 2026-08-23T15:58:41.000Z (ARR)
• Source URL: https://www.publickey1.jp/blog/26/rustbun_14nodejsplaywrightvitestcpu.html