Goeteia: Scheme-to-WebAssembly Compiler Explained
12 Jul 2026
What launched
A new open-source project called Goeteia compiles pure Scheme code directly to WebAssembly for execution in browsers and other WASM engines. The compiler is notable for being self-hosted — it's written in the same Scheme subset it compiles — and its source is published on GitHub at github.com/guenchi/Goeteia.
The technical details
According to the report, Goeteia's core numbers are:
- The compiler's WebAssembly file is approximately 70 KB gzipped
- Each compilation of Scheme source code takes approximately 15 milliseconds
- A 100 million iteration loop runs in approximately 150 milliseconds
Compiled Scheme modules run on any WebAssembly engine that supports garbage collection and tail calls — a list that currently includes Node 22+, current versions of Chrome, Firefox, and Safari, and wasmtime.
Bootstrapping the compiler from source requires Chez Scheme, but the checked-in compiler WebAssembly binary works without it — meaning most users won't need the Chez toolchain unless they're building from scratch.
What's missing
The report flags several open questions. There's no stated release date, versioning, or named maintainers. It's unclear which Scheme standard or subset (such as R7RS) Goeteia implements, and there's no comparison to other Scheme-to-WASM or general-purpose language-to-WASM compilers. Licensing terms aren't specified, and the methodology behind the 70 KB size, 15ms compile time, and 150ms loop benchmark — including hardware and comparison baseline — isn't disclosed. There's also no data yet on adoption, community size, or production use cases.
Risks to weigh
A few structural risks stand out:
- Browser/engine compatibility: Dependency on WebAssembly GC and tail-call support may limit compatibility with older browsers or engines that lack these features.
- Contributor friction: Reliance on Chez Scheme for bootstrapping from source could slow down contributors who don't already have that toolchain installed.
- Unclear longevity: As a new, niche tool, long-term maintenance and community support levels remain unclear.
Why founders should care
For founders building browser-based or WASM-targeted products, Goeteia is likely worth a bookmark rather than an immediate integration. The small footprint (~70 KB gzipped) and fast reported compile/execution times suggest it could suit performance- or size-sensitive applications, but the lack of published benchmark methodology limits how much confidence to place in those numbers. Support across Node 22+, Chrome, Firefox, Safari, and wasmtime broadens potential deployment targets, which may matter if you're evaluating polyglot or WASM-first architectures. At the same time, the fact that it's self-hosted and open-source but lacking clear versioning or adoption data suggests this is more likely an experimental or maturing project than a production-ready dependency today. Teams considering it should verify target browser and engine support for WebAssembly GC and tail calls before committing, and should treat the current benchmarks as directional rather than definitive.
Bottom line
Goeteia adds a lightweight, self-hosted option to the small but growing space of language-to-WebAssembly compilers. Its real value to founders will depend on details not yet public — licensing, language coverage, and community traction — making it a project to watch rather than adopt outright.