· 2 min read ·

TypeScript 7 Is Going Native and It's About Time

Source: typescript

If you’ve ever stared at a TypeScript compile step grinding through a large monorepo and thought “this shouldn’t be this slow,” you’re going to want to pay attention to what the TypeScript team announced late last year.

Project Corsa — now officially being called TypeScript 7.0 — is a full port of the TypeScript compiler and language service to native code. The December 2025 progress update confirms the team has made significant strides, and honestly, the ambition here is remarkable.

The Problem With TypeScript Compiling TypeScript

Here’s something that’s always been a little funny: the TypeScript compiler is itself written in TypeScript, running on Node.js, which runs on V8. You’re paying a runtime tax every single time you compile. For small projects it’s fine. For a large codebase — say, a Discord bot with a complex event system, dozens of modules, and strict mode on everything — type-checking can become a real friction point in your dev loop.

The current compiler has always been architecturally constrained by JavaScript’s single-threaded execution model. Even with worker threads, you’re fighting the runtime rather than working with it.

What Native Means in Practice

Porting to native code means the compiler can:

  • Actually use multiple CPU cores for parallel compilation
  • Manage memory directly, avoiding V8’s garbage collector overhead
  • Run faster raw execution, without the JIT warmup costs

For IDE integrations — the language server that powers autocomplete, go-to-definition, and inline errors — this is potentially transformational. Right now, large files can make the language service feel sluggish. A native implementation could make that feel instant.

My Take

I’ve been following this since the initial announcement earlier in 2025, and I’ll admit I was skeptical about the timeline. Rewriting a compiler of this complexity is not a weekend project. TypeScript’s type system is famously expressive — and famously complex to implement correctly. The fact that they’re showing real progress by December is encouraging.

The thing that excites me most isn’t raw compile speed, though that’s a clear win. It’s what parallelism unlocks. Type-checking individual files or modules simultaneously, background re-checking on save without blocking the editor — these are quality-of-life improvements that compound over a workday.

For anyone building tooling on top of TypeScript — linters, bundlers, code generators — a faster and more parallel language service API is also a huge deal. The whole ecosystem gets faster.

What to Watch For

The TypeScript team is being appropriately cautious. A native rewrite has to be a drop-in replacement — it needs to type-check existing code identically, support the same compiler flags, and integrate with the same tooling. Any subtle behavioral difference becomes a bug report. Getting that right under a .tsconfig you’ve had for years is a high bar.

Typescript 7.0 doesn’t have a release date yet, but the momentum looks real. If you want to follow along, the Microsoft DevBlog is the right place to watch. This is shaping up to be one of the more meaningful infrastructure improvements in the JavaScript ecosystem in years.

Was this interesting?