Swiftpack.co - google/swift-benchmark as Swift Package

Swiftpack.co is a collection of thousands of indexed Swift packages. Search packages.
See all packages published by google.
google/swift-benchmark 0.1.2
A swift library to benchmark code snippets.
⭐️ 785
🕓 13 weeks ago
.package(url: "https://github.com/google/swift-benchmark.git", from: "0.1.2")


A Swift library for benchmarking code snippets, similar to google/benchmark.


import Benchmark

benchmark("add string reserved capacity") {
    var x: String = ""
    for _ in 1...1000 {
        x += "hi"


At runtime, you can filter which benchmarks to run by using the --filter command line flag. For more details on what options are available, pass either the -h or --help command line flags.


$ swift run -c release BenchmarkMinimalExample --help
USAGE: benchmark-command [--allow-debug-build] [--filter <filter>] [--filter-not <filter-not>] [--iterations <iterations>] [--warmup-iterations <warmup-iterations>] [--min-time <min-time>] [--max-iterations <max-iterations>] [--time-unit <time-unit>] [--inverse-time-unit <inverse-time-unit>] [--columns <columns>] [--format <format>] [--quiet]

  --allow-debug-build     Overrides check to verify optimized build.
  --filter <filter>       Run only benchmarks whose names match the regular expression.
  --filter-not <filter-not>
                          Exclude benchmarks whose names match the regular expression.
  --iterations <iterations>
                          Number of iterations to run.
  --warmup-iterations <warmup-iterations>
                          Number of warm-up iterations to run.
  --min-time <min-time>   Minimal time to run when automatically detecting number iterations.
  --max-iterations <max-iterations>
                          Maximum number of iterations to run when automatically detecting number iterations.
  --time-unit <time-unit> Time unit used to report the timing results.
  --inverse-time-unit <inverse-time-unit>
                          Inverse time unit used to report throughput results.
  --columns <columns>     Comma-separated list of column names to show.
  --format <format>       Output format (valid values are: json, csv, console, none).
  --quiet                 Only print final benchmark results.
  -h, --help              Show help information.

$ swift run -c release BenchmarkMinimalExample
running add string no capacity... done! (1832.52 ms)
running add string reserved capacity... done! (1813.96 ms)

name                         time     std        iterations
add string no capacity       37435 ns ±   6.22 %      37196
add string reserved capacity 37022 ns ±   1.75 %      37749

For more examples, see Sources/BenchmarkMinimalExample and Sources/BenchmarkSuiteExample.


Add this library as a SwiftPM dependency:

let package = Package(
    name: ... ,
    products: [
        .executable(name: "Benchmarks", targets: ["Benchmarks"])
    dependencies: [
      .package(url: "https://github.com/google/swift-benchmark", from: "0.1.0")
    targets: [
            name: "Benchmarks",
            dependencies: [.product(name: "Benchmark", package: "swift-benchmark")]


The project is in an early stage and offers only a basic set of benchmarking utilities. Feel free to file issues and feature requests to help us prioritize what to do next.


Please see LICENSE for details.


Please see CONTRIBUTING.md for details.


Stars: 785
Last commit: 13 weeks ago
jonrohan Something's broken? Yell at me @ptrpavlik. Praise and feedback (and money) is also welcome.

Release Notes

13 weeks ago

Minor release that bumps version of swift-argument parser dependency, and adds a benchmark comparison script.

Changes since last release:

  • Add script to compare benchmark results (#85)
  • Clarify what's really needed for SPM (#91)
  • CI: Simplify the Windows CI rules (#93)
  • Bump swift-argument-parser to stable (#94)

Swiftpack is being maintained by Petr Pavlik | @ptrpavlik | @swiftpackco | API | Analytics