readme.md

Interpreted Languages Benchmark

Benchmark for interpreted languages.

The values for Rust, V, Swift, D and Haskell are not really representative as they are compiled on the first run and Haskell unfortunately does not even cache the compiled binary.

Languages

• Bash
• Bun
• D
• Dart
• Dash
• Elixir
• Elvish
• F#
• Fish
• Guile
• Haskell
• Java
• JavaScript via
  • Node.js
  • Deno
  • Bun
• Julia
• Ksh
• Lua
• Lua
• Nickel
• Nix Language
• Nushell
• OCaml
• Osh
• Perl
• PHP
• TypeScript via
  • Deno
  • Bun
• Python
• Python
• R
• Racket
• Roc.roc
• Ruby
• Scala
• Swift
• Typst
• V

Workarounds

• Typst
  Can only output a JSON string. Use … | jq -r to remove the quotes.

Result

Check out the workflow runs for the latest benchmark results.

Related

• ffi-overhead - Comparing the C FFI overhead of various programming languages
• gradbench - Benchmarks for differentiable programming across languages and domains
• jinyus/related_post_gen - Data Processing benchmark
• plb2 - A programming language benchmark
• Programming-Language-Benchmarks
• script-bench-rs - Rust embedded scripting languages benchmark