tsrs explained: can a Rust port make TypeScript checks 8x faster?

tsrs is an experimental Rust port of the TypeScript compiler, announced by Cloudflare on 20 October 2026. Cloudflare and VoidZero are building it with Max Schwenk of Owner.com, in collaboration with the TypeScript team, and it is incubated in the Oxc toolchain.

Early benchmarks show it up to 8× faster, with about 70% less memory, than TypeScript 7.0 using the same number of type-check workers. These are the authors' benchmarks, and the hardware is not listed.

Sources checked on 26 October 2026. Features, prices and availability can change; confirm the linked primary sources before acting.

What tsrs is, and what it is not

It is a faithful port, not a new language or a fork. The authors say the TypeScript team keeps full control of language design.

The goals are shorter waits for type checks and lower memory use, so full development environments can run in more places, including browsers. A WebAssembly package, @tsrs/wasm, is available. The authors argue Rust suits WebAssembly better than Go because Go's Wasm builds must include its garbage collector.

How tsrs compares with TypeScript 7

TypeScript 7 is Microsoft's official Go port; tsrs is an independent Rust port. The table compares what each project claims.

	TypeScript 7 (Go)	tsrs (Rust)
Status	Official compiler	Experimental, no stable release
Speed claim	7.1 beta up to 2.5× faster than 7.0	Up to 8× faster than 7.0 at equal workers
Memory claim	Not stated	About 70% less than 7.0
Output parity	Reference	Known differences, such as diagnostic order

Some tsrs optimisations have already been contributed back to TypeScript 7.1, so the gap may narrow. By default tsrs uses an adaptive configuration that uses more memory and runs faster still, so match settings when you compare.

Test tsrs against a real monorepo

Run it beside your current checker; do not replace it yet. A side-by-side run shows speed and correctness together.

1. On a branch, install the tsrs package from npm and run npx tsrs at the project root, following the repository README for options.
2. Time cold and warm runs of your current check and of tsrs with the same worker count.
3. Record peak memory with /usr/bin/time -v on Linux or /usr/bin/time -l on macOS.
4. Diff the diagnostics after sorting them, because print order differs.
5. Repeat on your largest package, not a small example.

Where tsrs could matter: browser and AI coding environments

Memory, not only speed, decides whether a type checker fits in a sandbox. Browser-based and AI coding environments run many isolated workspaces at once.

A smaller footprint could let an agent type-check its own changes inside a sandbox before proposing them. That depends on output parity, though. Until a stable release, keep the official compiler as the gate in CI.

Our verdict: tsrs is worth a benchmark afternoon for large monorepos, not a CI replacement. Related: upgrading to TypeScript 7. If slow builds are hurting your team, our React and Next.js frontend team can audit them.

Sources and further reading

• A Rust port of TypeScript — Cloudflare
• tsrs repository — Oxc on GitHub

Image credits

Explanatory diagrams: Cloudsoft Zone, CC BY-SA 4.0. Featured photograph: HaeB, CC BY-SA 4.0, via Wikimedia Commons. RustCodeOnScreen: Slashme, CC0, via Wikimedia Commons.