What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Yes—an AI coding-agent experiment produced a working Rust port of TypeScript 7’s compiler, type checker and language server for about $24,047 in reported API spend. The project, called ts-rust or tsc-rs, is not a finished replacement for the official compiler. It is an early, machine-generated implementation whose compatibility and speed claims come from its own 2026 repository tests and benchmarks.
The result is notable because an earlier attempt consumed more than $420,000 in OpenAI-model API tokens, generated over 1.3 million lines of Rust and stalled at roughly 84% compatibility. The successful run started over and reached a working version in about 10 hours during a two-week Claude Opus 5.5 effort.
As an Amazon Associate I earn from qualifying purchases.
What ts-rs actually is
TypeScript 7’s new native toolchain is implemented in Go. Theo Browne’s ts-rs project attempts to reproduce that implementation in Rust, including the compiler, type checker and language server. It is also published as the npm package tsc-rs and as source software through GitHub.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteThe goal is broader than translating syntax. The Rust program aims to retain the command-line behavior, language-server behavior and API surface of the Go implementation. That makes it a potential drop-in candidate for editor tooling and build systems, provided its output remains compatible with the particular TypeScript development revision it targets.
#1 Best Overall
This is an experiment rather than an official Microsoft release. The repository is pinned to a specific TypeScript development revision, so comparisons are meaningful only when the upstream revision, compiler options and input project match.
How the $420,000 failure led to the $24,047 result
| Attempt | Reported spend and duration | Result |
|---|---|---|
| Earlier OpenAI-model attempts | More than $420,000 in API-priced tokens, according to Theo Browne’s 2026 account | More than 1.3 million lines of Rust; work stalled at about 84% compatibility |
| Successful Claude Opus 5.5 run | About $24,047 in API spend over two weeks, according to Browne’s 2026 account | Started from scratch and produced a working version in about 10 hours |
These are the maintainer’s reported token bills, not a standardized cost-per-line experiment. They include model API usage and do not establish what another team would pay for a similar port. The contrast does show that more generated code and more spending did not guarantee completion: restarting with a different agent workflow produced a better result at a much lower reported cost.
Rank #2
What the project says it has achieved
Large-scale test coverage
The 2026 project README reports 184,773 ported Go tests passing. That is a project-reported count, not an independent certification. Passing tests from the Go implementation is useful evidence that many language and compiler behaviors were reproduced, but it does not cover every possible program or operating environment.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Real-project diagnostics
The README also reports matching diagnostics for named open-source projects including TanStack Query core and Hono. It reports matching language-server results and API-oracle results as well. “Matching” here means the project’s comparison procedures found equivalent results for the cases tested; it is not a promise that every editor action or API call will behave identically.
Rank #3
How fast is tsc-rs compared with TypeScript 7?
The repository’s dated benchmark run was measured on October 13, 2026 UTC using Blacksmith 16-vCPU Ubuntu runners. The numbers below are seconds for the T3 Code project under the stated diagnostics configuration.
| Implementation | With Effect diagnostics | Without Effect diagnostics |
|---|---|---|
tsc-rs --fast |
9.77 seconds | 6.91 seconds |
tsc-rs |
14.30 seconds | 7.95 seconds |
tsc 7 |
42.55 seconds | 19.45 seconds |
tsc 6 |
191.00 seconds | Not stated in the README’s reported table |
On the Effect-diagnostics run, the fast Rust mode completed in roughly one-fifth of the TypeScript 7 time. Disabling those diagnostics changed every result, which is why a single speed claim can be misleading. The README’s broader comparison across six open-source applications reports version 0.6.0 of tsc-rs as 2.10× faster than tsc 7 on geometric mean. That figure applies only to the repository’s versions, flags, hardware and workload selection; it is not a universal speedup for every project.
What the results do—and do not—prove
- They do show: an AI-driven workflow can generate a substantial Rust implementation of a production-scale compiler and reach high reported test and compatibility coverage on selected cases.
- They do show: the Rust build can be materially faster than the compared TypeScript 7 and TypeScript 6 runs under the published benchmark conditions.
- They do not show: complete language compatibility, identical behavior on every codebase, stable performance across hardware, or long-term maintenance capacity.
- They do not show: that the reported $24,047 is a repeatable budget for another compiler port.
Is the AI-written Rust compiler safe to use?
The safest answer is to treat it as an experimental tool that requires independent validation before it enters a production build or editor fleet. The maintainer explicitly warns that the release is early, documents known edge cases and says the generated source has not been manually read.
“I’ve never read a line of this code.” — Theo Browne, ts-rust maintainer, project README (2026)
That statement matters even alongside a large passing-test count. Compiler correctness includes undefined-behavior boundaries, resource exhaustion, platform-specific failures, error recovery and interactions that a test suite may not exercise. Generated code can also be difficult to maintain when upstream TypeScript changes.
Checks to perform before adoption
- Pin the exact TypeScript development revision used by the repository and compare it with the revision in your own toolchain.
- Run both compilers against representative projects, including generated code, strict-mode code and files that rely on advanced type-system behavior.
- Diff diagnostics, emitted output, exit codes and language-server responses rather than checking only whether compilation succeeds.
- Keep the official compiler as a fallback in continuous integration and test the Rust binary on every supported operating system and CPU architecture.
- Monitor the project for fixes to its documented edge cases and for changes in its generated source or benchmark methodology.
Tools such as Bun may be faster in some situations while producing different diagnostics, so speed alone is not a correctness test.
Why Rust could matter beyond raw speed
The Register has pointed to a possible systems advantage: Rust has no garbage collector, which could make WebAssembly deployment inside a V8-based agent environment easier than embedding a Go runtime. That is a proposed use case, not a demonstrated production deployment of tsc-rs. The current evidence concerns compiler behavior and benchmark runs, not a finished WebAssembly integration.
How to interpret the experiment’s real significance
The headline is not that AI has replaced compiler engineering for $24,000. The stronger, narrower conclusion is that an agentic coding workflow can now attempt a difficult, testable systems port at a scale that previously required a large human team—and can sometimes recover from a failed approach by restarting with a better strategy.
For users, tsc-rs is best viewed as a promising drop-in candidate for controlled experiments, performance investigations and further engineering. For maintainers, its test count, project comparisons and dated benchmarks are useful starting evidence, while the pinned revision, known edge cases and absence of manual code review remain significant adoption risks.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
