diff --git a/.github/actionlint.yaml b/.github/actionlint.yaml
index 3c69aef3..b0de1ec3 100644
--- a/.github/actionlint.yaml
+++ b/.github/actionlint.yaml
@@ -1,7 +1,2 @@
 # Configuration for the actionlint tool, which we run via prek
 # to verify the correctness of the syntax in our GitHub Actions workflows.
-
-self-hosted-runner:
-  # Various runners we use that aren't recognized out-of-the-box by actionlint:
-  labels:
-    - codspeed-macro
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 7917a81c..8f695238 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -243,10 +243,10 @@ jobs:
       - name: Build docs
         run: uv run --isolated --only-group docs zensical build
 
-  benchmarks-walltime:
-    name: "walltime benchmarks"
+  benchmarks-instrumented:
+    name: "instrumented benchmarks"
 
-    runs-on: ubuntu-latest # should be codspeed-macro
+    runs-on: ubuntu-latest
 
     needs: determine_changes
     if: ${{ (needs.determine_changes.outputs.code == 'true' || github.ref == 'refs/heads/main') }}
@@ -283,7 +283,7 @@ jobs:
       - name: Build benchmarks
         env:
           PYO3_PYTHON: ${{ steps.setup-python.outputs.python-path }}
-        run: cargo codspeed build -m walltime
+        run: cargo codspeed build
 
       - name: Run benchmarks
         uses: CodSpeedHQ/action@972e3437949c89e1357ebd1a2dbc852fcbc57245 # v4.5.1
@@ -291,5 +291,5 @@ jobs:
           PYO3_PYTHON: ${{ steps.setup-python.outputs.python-path }}
         with:
           token: ${{ secrets.CODSPEED_TOKEN }}
-          run: cargo codspeed run --bench karva_benchmark_walltime
-          mode: walltime
+          run: cargo codspeed run --bench karva_benchmark
+          mode: simulation
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index fa0e215b..e4cd871c 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -64,7 +64,7 @@ Infrastructure and tooling:
 - `karva_python` — the PyO3 `cdylib` that produces the Python wheel and wraps both `karva` and `karva_worker`.
 - `karva_macros` — procedural macros.
 - `karva_dev` — dev tools such as CLI reference generation.
-- `karva_benchmark` — wall-time benchmark that runs `karva test` against a pinned snapshot of `karva-benchmark-1`.
+- `karva_benchmark` — benchmark that runs `karva test` against a pinned snapshot of `karva-benchmark-1`.
 
 ### Prerequisites
 
diff --git a/Cargo.lock b/Cargo.lock
index 69f80a82..e90653a8 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1108,7 +1108,6 @@ dependencies = [
  "karva_metadata",
  "karva_project",
  "karva_runner",
- "karva_static",
  "ruff_python_ast",
  "serde",
  "serde_json",
diff --git a/crates/karva_benchmark/Cargo.toml b/crates/karva_benchmark/Cargo.toml
index 33d8ea0c..4cbf6d27 100644
--- a/crates/karva_benchmark/Cargo.toml
+++ b/crates/karva_benchmark/Cargo.toml
@@ -20,13 +20,12 @@ karva_logging = { workspace = true }
 karva_metadata = { workspace = true }
 karva_project = { workspace = true }
 karva_runner = { workspace = true }
-karva_static = { workspace = true }
 ruff_python_ast = { workspace = true }
 serde = { workspace = true }
 serde_json = { workspace = true }
 
 [[bench]]
-name = "karva_benchmark_walltime"
+name = "karva_benchmark"
 harness = false
 
 [lints]
diff --git a/crates/karva_benchmark/benches/karva_benchmark_walltime.rs b/crates/karva_benchmark/benches/karva_benchmark.rs
similarity index 100%
rename from crates/karva_benchmark/benches/karva_benchmark_walltime.rs
rename to crates/karva_benchmark/benches/karva_benchmark.rs
diff --git a/crates/karva_benchmark/src/lib.rs b/crates/karva_benchmark/src/lib.rs
index a4038d9e..5ea8c32a 100644
--- a/crates/karva_benchmark/src/lib.rs
+++ b/crates/karva_benchmark/src/lib.rs
@@ -1,4 +1,4 @@
-//! Wall-time benchmark for karva.
+//! Benchmark for karva.
 //!
 //! Clones a fixed snapshot of <https://github.com/karva-dev/karva-benchmark-1>
 //! into `target/benchmark_cache/`, installs its dependencies via uv, and runs
@@ -36,7 +36,7 @@ pub fn setup_project() -> Project {
 /// Run karva tests against the prepared project once.
 pub fn run_karva(project: &Project) {
     let config = karva_runner::ParallelTestConfig {
-        num_workers: karva_static::max_parallelism().get(),
+        num_workers: 2,
         no_cache: false,
         create_ctrlc_handler: false,
         last_failed: false,