
Bun 1.4 is out, rewritten in Rust. It fixes over 2,900 issues and adds 1,517 Node.js tests.
Idle CPU use drops 5x, memory use falls 35%, and startup is 50% faster on Linux.
Playwright, vitest, and OpenTelemetry now work with it.
What happened
Bun, the server-side JavaScript runtime, has released version 1.4, the first version rewritten from Zig to Rust using Claude Code. The rewrite fixed over 2,900 GitHub issues and added 1,517 tests from the Node.js test suite.
Why it matters
The Rust rewrite brings major performance and compatibility improvements. Idle CPU usage is reduced by 5x, memory usage is down by up to 35%, and startup is up to 50% faster on Linux. Frameworks like Playwright, vitest, and OpenTelemetry now work, and a memory leak in Next.js App Router SSR was fixed.
What to watch
The performance gains are notable, with regex processing up to 281x faster compared with Node.js and startup time improved to 4.8 milliseconds. Bun states full Node.js compatibility is not yet achieved, so future versions may close that gap.
Ask the AI about this article →
Bun 1.4 is the first version to be rewritten in Rust, a move that directly addresses performance and stability. The rewrite was done with Claude Code over 11 days, a detail highlighted by the Bun team. The resulting fixes and optimizations are substantial, from a 5x reduction in idle CPU usage to a 35% cut in memory use. Startup time on Linux has also improved by up to 50%, bringing it to 4.8 milliseconds. These gains stem from the Rust port and memory-related stabilizations. The version also marks a significant push toward Node.js compatibility, adding over 1,500 tests from the official suite, which enabled frameworks like Playwright and vitest to function. However, Bun notes that full Node.js compatibility is still a work in progress. The memory leak fix for Next.js App Router SSR, caused by a GC retention cycle when fetch() is called, is another key improvement that could benefit developers using that framework.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
Ask AI anything about this article. Q&As are published on this page for other readers too.
Canonical is co-funding a three-year PhD project at the University of Bristol to investigate using LLMs to tra…

In 9 days from Aug 10, Meta (Muse Glimmer), NVIDIA (Nemotron 3.5 Lightning), and Alibaba Cloud (Qwen3.8-27B) r…

An anonymous provider has been offering free access to a frontier-class coding model called Ox Alpha on OpenRo…
Peter Steinberger built OpenClaw, an open-source agent harness that lets users control AI agents through Whats…

Airbnb's AI customer-service assistant now resolves nearly 45% of user inquiries without a human agent, up fro…

AT&T cut the costs of coding and some other advanced AI tasks by as much as 56% by using LiteLLM model routers…
