cargo bench: Usage, Options & Common CI Errors

Compile and run benchmark targets.

cargo bench builds your crate with optimizations and runs benchmark targets. The built-in #[bench] harness is unstable, so most projects benchmark with Criterion on stable Rust.

What it does

Compiles in the bench profile (optimized) and executes registered benchmarks. Arguments after -- pass to the bench harness, mirroring cargo test.

Common usage

Terminal

cargo bench                       # run all benchmarks
cargo bench parse                 # filter by name
cargo bench --no-run              # build benches only
cargo bench -- --save-baseline ci # Criterion baseline

Common CI error: bench attribute needs nightly

Using #[bench] / test::Bencher on stable fails with "error[E0658]: use of unstable library feature test." Fix: either run on the nightly toolchain, or switch to Criterion (a harness = false bench target in Cargo.toml) which runs on stable.

Options

Flag	Effect
--no-run	Compile benches without running
--bench <name>	Run a specific bench target
-- <args>	Pass args to the harness

Related guides

cargo test: Usage, Options & Common CI ErrorsHow cargo test builds and runs unit, integration, and doc tests, the flags for filtering and CI, and the comm… cargo build: Usage, Options & Common CI ErrorsWhat cargo build does, the flags you use most in CI (--release, --locked, --target), and how to fix the depen… Rust "cargo test failed" - Find the Failing Test in CIFix a failing "cargo test" in CI -- a real assertion failure, a panic, or a flaky test. Read the "test result…