diff --git a/mlx/backend/cpu/primitives.cpp b/mlx/backend/cpu/primitives.cpp
index 3299b74717..f70a2740df 100644
--- a/mlx/backend/cpu/primitives.cpp
+++ b/mlx/backend/cpu/primitives.cpp
@@ -380,6 +380,10 @@ void View::eval_cpu(const std::vector<array>& inputs, array& out) {
       strides[i] *= ibytes;
       strides[i] /= obytes;
     }
+    // Last axis can have any stride
+    if (obytes < ibytes) {
+      strides.back() = 1;
+    }
     out.copy_shared_buffer(
         in, strides, in.flags(), in.data_size() * ibytes / obytes);
   } else {
diff --git a/mlx/backend/gpu/primitives.cpp b/mlx/backend/gpu/primitives.cpp
index 0bc19c8171..72da082c67 100644
--- a/mlx/backend/gpu/primitives.cpp
+++ b/mlx/backend/gpu/primitives.cpp
@@ -255,6 +255,10 @@ void View::eval_gpu(const std::vector<array>& inputs, array& out) {
       strides[i] *= ibytes;
       strides[i] /= obytes;
     }
+    // Last axis can have any stride
+    if (obytes < ibytes) {
+      strides.back() = 1;
+    }
     out.copy_shared_buffer(
         in, strides, in.flags(), in.data_size() * ibytes / obytes);
   } else {
diff --git a/python/tests/test_ops.py b/python/tests/test_ops.py
index b3589580e5..f26541030f 100644
--- a/python/tests/test_ops.py
+++ b/python/tests/test_ops.py
@@ -4154,6 +4154,11 @@ def test_view(self):
             a_out = out.view(mx.int32)
             self.assertTrue(mx.array_equal(a_out, a, equal_nan=True))
 
+        # Size one last axis
+        a = mx.arange(4, dtype=mx.int32).reshape(1, 4).T
+        expected = np.arange(4, dtype=np.int32).reshape(4, 1).view(np.int16)
+        self.assertTrue(np.array_equal(a.view(mx.int16), expected))
+
         # Byte offsets that are not a multiple of the new item size
         a = mx.arange(64, dtype=mx.uint8)
         a_np = np.arange(64, dtype=np.uint8)