diff --git a/Cargo.lock b/Cargo.lock
index 7f002ce298c..083d19187f0 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -4172,7 +4172,9 @@ name = "uu_uniq"
 version = "0.2.2"
 dependencies = [
  "clap",
+ "codspeed-divan-compat",
  "fluent",
+ "tempfile",
  "uucore",
 ]
 
diff --git a/src/uu/uniq/Cargo.toml b/src/uu/uniq/Cargo.toml
index 6b90eaad344..9504588bf8b 100644
--- a/src/uu/uniq/Cargo.toml
+++ b/src/uu/uniq/Cargo.toml
@@ -22,6 +22,11 @@ clap = { workspace = true }
 uucore = { workspace = true, features = ["parser"] }
 fluent = { workspace = true }
 
+[dev-dependencies]
+divan = { workspace = true }
+tempfile = { workspace = true }
+uucore = { workspace = true, features = ["benchmark", "parser"] }
+
 [[bin]]
 name = "uniq"
 path = "src/main.rs"
diff --git a/src/uu/uniq/benches/uniq_bench.rs b/src/uu/uniq/benches/uniq_bench.rs
new file mode 100644
index 00000000000..2b012e57d2c
--- /dev/null
+++ b/src/uu/uniq/benches/uniq_bench.rs
@@ -0,0 +1,113 @@
+// This file is part of the uutils coreutils package.
+//
+// For the full copyright and license information, please view the LICENSE
+// file that was distributed with this source code.
+
+use divan::{Bencher, black_box};
+use uu_uniq::uumain;
+use uucore::benchmark::{run_util_function, setup_test_file};
+
+/// Generate data with many consecutive duplicate lines
+/// This directly tests the core optimization of PR #8703 - avoiding allocations when comparing lines
+fn generate_duplicate_heavy_data(num_groups: usize, duplicates_per_group: usize) -> Vec<u8> {
+    let mut data = Vec::new();
+
+    for group in 0..num_groups {
+        // Generate a line with realistic content
+        let line = format!(
+            "Line content for group {group:06} with additional text to make it more realistic for testing performance\n"
+        );
+
+        // Repeat the line multiple times (this is what PR #8703 optimizes)
+        for _ in 0..duplicates_per_group {
+            data.extend_from_slice(line.as_bytes());
+        }
+    }
+
+    data
+}
+
+/// Benchmark 1: Heavy duplicates - the main optimization target
+/// Many consecutive duplicate lines that stress the line comparison optimization
+#[divan::bench(args = [10_000_000])]
+fn uniq_heavy_duplicates(bencher: Bencher, num_lines: usize) {
+    // Create 1000 groups with ~10,000 duplicates each
+    // This maximizes the benefit of PR #8703's optimization
+    let num_groups = 1000;
+    let duplicates_per_group = num_lines / num_groups;
+    let data = generate_duplicate_heavy_data(num_groups, duplicates_per_group);
+    let file_path = setup_test_file(&data);
+    let file_path_str = file_path.to_str().unwrap();
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &[file_path_str]));
+    });
+}
+
+/// Benchmark 2: Mixed duplicates with counting
+/// Tests the -c flag with a mix of duplicate groups
+#[divan::bench(args = [5_000_000])]
+fn uniq_with_count(bencher: Bencher, num_lines: usize) {
+    // Create more groups with fewer duplicates for varied counting
+    let num_groups = num_lines / 100;
+    let data = generate_duplicate_heavy_data(num_groups, 100);
+    let file_path = setup_test_file(&data);
+    let file_path_str = file_path.to_str().unwrap();
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &["-c", file_path_str]));
+    });
+}
+
+/// Benchmark 3: Case-insensitive comparison with duplicates
+/// Tests the -i flag which requires case folding during comparison
+#[divan::bench(args = [2_000_000])]
+fn uniq_case_insensitive(bencher: Bencher, num_lines: usize) {
+    let mut data = Vec::new();
+    let words = [
+        "Hello",
+        "WORLD",
+        "Testing",
+        "UNIQ",
+        "Benchmark",
+        "Performance",
+    ];
+
+    // Generate groups of case variations
+    for i in 0..num_lines {
+        let word = words[(i / 50) % words.len()];
+
+        // Create case variations that should be treated as duplicates with -i
+        let variation = match i % 4 {
+            0 => word.to_lowercase(),
+            1 => word.to_uppercase(),
+            2 => word.to_string(),
+            _ => {
+                // Mixed case
+                word.chars()
+                    .enumerate()
+                    .map(|(idx, c)| {
+                        if idx % 2 == 0 {
+                            c.to_lowercase().to_string()
+                        } else {
+                            c.to_uppercase().to_string()
+                        }
+                    })
+                    .collect()
+            }
+        };
+
+        data.extend_from_slice(format!("{variation}\n").as_bytes());
+    }
+
+    let file_path = setup_test_file(&data);
+    let file_path_str = file_path.to_str().unwrap();
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &["-i", file_path_str]));
+    });
+}
+
+fn main() {
+    divan::main();
+}