Js_of_ocaml is a compiler from OCaml bytecode programs to JavaScript. It makes it possible to write OCaml programs that run on Web browsers. It is easy to install as it works with an existing installation of OCaml, with no need to recompile any library. It comes with bindings for a large part of the browser APIs. According to our benchmarks, the generated programs runs typically faster than with the OCaml bytecode interpreter. We believe this compiler will prove much easier to maintain than a retargeted OCaml compiler, as the bytecode provides a very stable API.



  • Jbuilder
  • Cmdliner, cppo (for the compiler)
  • ppx_tools_versioned, ocaml-migrate-parsetree (for the js_of_ocaml base library)

See opam file at the root of the repository for version constraints.

Optional dependencies:

Install from opam

opam install js_of_ocaml

Build and install from source

make opam-installer js_of_ocaml-compiler opam-installer js_of_ocaml-lwt opam-installer js_of_ocaml


Your program must first be compiled using the OCaml bytecode compiler ocamlc. JavaScript bindings are provided by the js_of_ocaml package and the syntax extension by the js_of_ocaml-ppx package (or js_of_ocaml-camlp4)

      ocamlfind ocamlc -package js_of_ocaml -package js_of_ocaml-ppx \
          -linkpkg -o cubes.byte cubes.ml

Then, run the js_of_ocaml compiler to produce JavaScript code:

      js_of_ocaml cubes.byte

See also the plugin Ocamlbuild_js_of_ocaml.

See also Quickstart.

Supported features

Most of the OCaml standard library is supported. However,

  • Weak semantic cannot be implemented using JavaScript. A dummy implementation is available (use `+weak.js` option)
  • Most of Sys module is not supported.

Extra libraries distributed with Ocaml (such as Thread or Str) are not supported in general. However,

  • Bigarray: bigarray are supported using Typed Arrays
  • Num: supported using `+nat.js` option
  • Graphics: partially supported using canvas (see also js_of_ocaml-graphics)
  • Unix: time related functions are supported

Tail call is not optimized in general. However, mutually recursive functions are optimized:

  • self recursive functions (when the tail calls are the function itself) are compiled using a loop.
  • trampolines are used otherwise. More about tail call optimization.

Data representation differs from the usual one. Most notably, integers are 32 bits (rather than 31 bits or 63 bits), wich is their natural size in JavaScript, and floats are not boxed. As a consequence, marshalling, polymorphic comparison, and hashing functions can yield results different from usual:

  • marshalling of floats is not supported (unmarshalling works);
  • the polymorphic hash function will not give the same results on datastructures containing floats;
  • these functions may be more prone to stack overflow.