diff --git a/.github/workflows/bench.yml b/.github/workflows/bench.yml
index aef481e..ab034a5 100644
--- a/.github/workflows/bench.yml
+++ b/.github/workflows/bench.yml
@@ -1,55 +1,38 @@
+name: benches
+
 on:
   push:
     branches:
-      - main
+      - "main"
   pull_request:
-
-name: Benchmark
+  # `workflow_dispatch` allows CodSpeed to trigger backtest
+  # performance analysis in order to generate initial data.
+  workflow_dispatch:
 
 concurrency:
-  group: ${{ github.workflow }}-${{ github.ref_name }}-${{ github.event.pull_request.number || github.sha }}
+  group: ${{ github.workflow }}-${{ github.ref_name }}-${{ github.event.pull_request.number || github.sha }}-benches
   cancel-in-progress: true
 
 jobs:
-  benchmark:
-    name: Benchmark
+  benchmarks:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
-      - uses: actions-rs/toolchain@v1
+      - uses: actions/checkout@v4
+      - uses: dtolnay/rust-toolchain@stable
         with:
-          profile: minimal
-          toolchain: stable
-          override: true
-      - name: Cache cargo build
-        uses: Swatinem/rust-cache@v1
-      - name: Run benchmark
-        run: |
-          cargo bench --all-features --bench jieba_benchmark -- --output-format bencher | tee output.txt
+          components: rust-src
 
-      # Download previous benchmark result from cache (if exists)
-      - name: Download previous benchmark data
-        uses: actions/cache@v1
-        with:
-          path: ./cache
-          key: ${{ runner.os }}-benchmark
+      - uses: Swatinem/rust-cache@v2
+        continue-on-error: true
+
+      - name: Install cargo-codspeed
+        run: cargo install cargo-codspeed
+
+      - name: Build the benchmark target(s)
+        run: cargo codspeed build --all-features
 
-      # Run `github-action-benchmark` action
-      - name: Store benchmark result
-        uses: rhysd/github-action-benchmark@v1
+      - name: Run the benchmarks
+        uses: CodSpeedHQ/action@v2
         with:
-          name: jieba_benchmark
-          # What benchmark tool the output.txt came from
-          tool: 'cargo'
-          # Where the output from the benchmark tool is stored
-          output-file-path: output.txt
-          # # Where the previous data file is stored
-          # external-data-json-path: ./cache/benchmark-data.json
-          # Workflow will fail when an alert happens
-          fail-on-alert: true
-          # GitHub API token to make a commit comment
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          # Enable alert commit comment
-          comment-on-alert: true
-          alert-comment-cc-users: '@messense'
-          auto-push: true
+          run: cargo codspeed run --all-features
+          token: ${{ secrets.CODSPEED_TOKEN }}
diff --git a/Cargo.toml b/Cargo.toml
index b630d13..5392d02 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -14,7 +14,7 @@ edition = '2021'
 all-features = true
 
 [dev-dependencies]
-criterion = "0.5"
+codspeed-criterion-compat = "2.4.1"
 rand = "0.8"
 wasm-bindgen-test = "0.3.0"
 
diff --git a/benches/jieba_benchmark.rs b/benches/jieba_benchmark.rs
index 0988961..932b5b0 100644
--- a/benches/jieba_benchmark.rs
+++ b/benches/jieba_benchmark.rs
@@ -1,7 +1,4 @@
-#[macro_use]
-extern crate criterion;
-
-use criterion::{black_box, Criterion, Throughput};
+use codspeed_criterion_compat::{black_box, criterion_group, criterion_main, Criterion};
 use jieba_rs::{Jieba, KeywordExtract, TextRank, TfIdf, TokenizeMode};
 use lazy_static::lazy_static;