diff --git a/dynamic_programming/knapsack_memoized_clean.py b/dynamic_programming/knapsack_memoized_clean.py
new file mode 100644
index 000000000000..2b8e13d908aa
--- /dev/null
+++ b/dynamic_programming/knapsack_memoized_clean.py
@@ -0,0 +1,57 @@
+from functools import cache
+
+
+def knapsack_memoized(weights: list[int], values: list[int], capacity: int) -> int:
+    """
+    Solve 0/1 knapsack using memoization without global state.
+
+    Args:
+        weights: list of item weights
+        values: list of item values
+        capacity: maximum capacity of knapsack
+
+    Returns:
+        Maximum achievable value
+
+    >>> knapsack_memoized([1, 3, 4], [10, 20, 30], 4)
+    30
+    >>> knapsack_memoized([1, 2, 3], [10, 15, 40], 6)
+    65
+    >>> knapsack_memoized([], [], 5)
+    0
+    >>> knapsack_memoized([2, 3, 4], [4, 5, 6], 0)
+    0
+    """
+
+    if len(weights) != len(values):
+        raise ValueError("weights and values must be of same length")
+
+    n = len(weights)
+
+    @cache
+    def dp(index: int, remaining: int) -> int:
+        """
+        Recursive helper function for knapsack memoization.
+
+        Args:
+            index: current item index
+            remaining: remaining capacity of knapsack
+
+        Returns:
+            Maximum value achievable from current state
+
+        Note:
+            This function is internally tested via knapsack_memoized doctests.
+        """
+        if index == n or remaining == 0:
+            return 0
+
+        if weights[index] > remaining:
+            return dp(index + 1, remaining)
+
+        return max(
+            dp(index + 1, remaining),
+            values[index] + dp(index + 1, remaining - weights[index]),
+        )
+
+    return dp(0, capacity)