What's Next?

Now that you've run your first benchmark, here's where to go depending on what you want to do.

I want to compare multiple implementations

Use Suite mode: BenchmarkSuite - a fluent builder for running several benchmarks side-by-side with a comparison table. When your suite grows to need complex setup or dependency injection, graduate to Host mode.

→ Guide: BenchmarkSuite

I want a dedicated benchmark project with attribute-based discovery

Use Host mode: BenchmarkHost - mark methods with [Benchmark], point the host at your assembly, and control everything from the command line.

→ Guide: BenchmarkHost

I want richer terminal output

Add NBenchmark.Console and use ConsoleReporter. It produces a colour-coded table with a bar chart, significance indicators, and a footnote explaining the Error column.

→ ConsoleReporter

I want to save results to a file

Use MarkdownReporter, CsvReporter, or JsonReporter. They require only the core NBenchmark package and can be stacked with any other reporter.

→ Reporters

I want to tune the measurement settings

Change the number of iterations, warmup iterations, outlier mode, or confidence level.

→ Configuration

I want to understand the statistics

A full technical explanation of how every number in the output is calculated.

→ Advanced: Statistics

I want to run benchmarks from the command line without recompiling

Use BenchmarkHost, which parses CLI arguments automatically. You can filter benchmarks, change reporter, set output directories, and more.

→ CLI Reference

My benchmark class needs dependencies (a repository, DbContext, logger, etc.)

Add the optional NBenchmark.DependencyInjection companion package and use UseDependencyInjection<T> - your benchmark class can then take constructor dependencies that the container resolves.

→ Dependency Injection guide