diff --git a/src/core/include/openvino/op/strided_slice.hpp b/src/core/include/openvino/op/strided_slice.hpp
index 5535e292580..c4b0394fa26 100644
--- a/src/core/include/openvino/op/strided_slice.hpp
+++ b/src/core/include/openvino/op/strided_slice.hpp
@@ -112,14 +112,14 @@ public:
     bool evaluate(const HostTensorVector& output_values, const HostTensorVector& input_values) const override;
     OPENVINO_SUPPRESS_DEPRECATED_END
     bool has_evaluate() const override;
-    OPENVINO_SUPPRESS_DEPRECATED_START
-    OPENVINO_SUPPRESS_DEPRECATED_END
     bool evaluate_lower(TensorVector& outputs) const override;
     bool evaluate_upper(TensorVector& outputs) const override;
     bool evaluate_label(TensorLabelVector& output_labels) const override;
+    bool constant_fold(OutputVector& output_values, const OutputVector& inputs_values) override;
 
 private:
     AxisSet convert_mask_to_axis_set(const std::vector<int64_t>& mask) const;
+    bool indicies_input_has_and_set_bounds(const size_t port, const std::vector<int64_t>& masks) const;
 
     std::vector<int64_t> m_begin_mask;
     std::vector<int64_t> m_end_mask;
diff --git a/src/core/src/bound_evaluate.cpp b/src/core/src/bound_evaluate.cpp
index 2aeeea2b2ab..fb07ddaed06 100644
--- a/src/core/src/bound_evaluate.cpp
+++ b/src/core/src/bound_evaluate.cpp
@@ -49,7 +49,7 @@ bool are_same_tensor(const ov::Tensor& lhs, const ov::Tensor& rhs) {
            (lhs.data() == rhs.data());
 }
 
-bool are_equal(const ov::Tensor& lhs, const ov::Tensor& rhs, size_t element_limit = 10) {
+bool are_equal(const ov::Tensor& lhs, const ov::Tensor& rhs) {
     if (!lhs || !rhs) {
         return false;
     }
@@ -59,7 +59,7 @@ bool are_equal(const ov::Tensor& lhs, const ov::Tensor& rhs, size_t element_limi
     const auto& lhs_et = lhs.get_element_type();
     const auto& rhs_et = rhs.get_element_type();
 
-    auto are_eq = (lhs_et == rhs_et) && (lhs_shape == rhs_shape) && shape_size(lhs_shape) <= element_limit;
+    auto are_eq = (lhs_et == rhs_et) && (lhs_shape == rhs_shape);
 
     if (are_eq) {
         are_eq = memcmp(lhs.data(), rhs.data(), lhs.get_byte_size()) == 0;
@@ -317,10 +317,14 @@ std::pair<ov::Tensor, ov::Tensor> ov::evaluate_both_bounds(const Output<Node>& o
             bool labels_evaluated = node->evaluate_label(output_labels);
             for (size_t i = 0; i < node->get_output_size(); ++i) {
                 auto& out_tensor = node->get_output_tensor(i);
+
                 out_tensor.set_lower_value(outputs_lower[i]);
-                out_tensor.set_upper_value(outputs_upper[i]);
-                if (same_inputs || are_equal(outputs_lower[i], outputs_upper[i]))
+                if (same_inputs || are_equal(outputs_lower[i], outputs_upper[i])) {
                     out_tensor.set_upper_value(outputs_lower[i]);
+                } else {
+                    out_tensor.set_upper_value(outputs_upper[i]);
+                }
+
                 if (labels_evaluated)
                     node->get_output_tensor(i).set_value_label(output_labels[i]);
             }
diff --git a/src/core/src/op/strided_slice.cpp b/src/core/src/op/strided_slice.cpp
index 1ba48d885df..97981412aa5 100644
--- a/src/core/src/op/strided_slice.cpp
+++ b/src/core/src/op/strided_slice.cpp
@@ -18,6 +18,7 @@
 #include "ngraph/slice_plan.hpp"
 #include "ngraph/type/element_type_traits.hpp"
 #include "ngraph/util.hpp"
+#include "openvino/core/validation_util.hpp"
 #include "openvino/op/util/precision_sensitive_attribute.hpp"
 #include "strided_slice_shape_inference.hpp"
 
@@ -65,6 +66,17 @@ shared_ptr<Node> calculate_default_strides(const Output<Node>& begin, const Outp
 
     return op::Constant::create(element::i64, ov::Shape{strides_length}, vector<int64_t>(strides_length, 1));
 }
+
+/**
+ * @brief Check if all indicies in 1-D input shape are ignored by masks.
+ *
+ * @param shape        Indicies shape (assume compatible 1-D shape).
+ * @param ignored_mask Axis set of ignored indicies.
+ * @return True if all ignored other wise false.
+ */
+bool all_indicies_ignored(const ov::PartialShape& shape, const AxisSet& ignore_mask) {
+    return shape.rank().is_static() && ov::cmp::le(shape[0].get_interval().get_max_val(), ignore_mask.size());
+}
 }  // namespace
 
 op::v1::StridedSlice::StridedSlice(const Output<Node>& data,
@@ -237,27 +249,75 @@ bool op::v1::StridedSlice::has_evaluate() const {
     return get_input_size() == 4;
 }
 
-namespace {
-bool strided_slice_input_check(const ov::Node* node) {
-    if (!node->get_input_tensor(1).has_and_set_bound() || !node->get_input_tensor(2).has_and_set_bound() ||
-        !node->get_input_tensor(3).has_and_set_bound())
-        return false;
-    return true;
+bool op::v1::StridedSlice::indicies_input_has_and_set_bounds(const size_t port,
+                                                             const std::vector<int64_t>& masks) const {
+    const auto& lb_t = get_input_tensor(port).get_lower_value();
+    const auto& ub_t = get_input_tensor(port).get_upper_value();
+
+    const auto mask_set = convert_mask_to_axis_set(masks);
+    bool valid_bounds = all_indicies_ignored(get_input_partial_shape(port), mask_set);
+
+    if (!valid_bounds && lb_t && ub_t) {
+        using TCast = int64_t;
+        constexpr auto i64_cast = ov::util::Cast<TCast>();
+        const auto lb = ov::get_tensor_data_as<TCast>(lb_t, i64_cast);
+        const auto ub = ov::get_tensor_data_as<TCast>(ub_t, i64_cast);
+
+        size_t axis = 0;
+        valid_bounds =
+            std::equal(lb.cbegin(), lb.cend(), ub.cbegin(), [&axis, &mask_set](TCast lhs, TCast rhs) -> bool {
+                return mask_set.count(axis++) || lhs == rhs;
+            });
+    }
+
+    return valid_bounds;
 }
-}  // namespace
 
 bool op::v1::StridedSlice::evaluate_lower(ov::TensorVector& output_values) const {
-    return strided_slice_input_check(this) && default_lower_bound_evaluator(this, output_values);
+    return indicies_input_has_and_set_bounds(1, get_begin_mask()) &&
+           indicies_input_has_and_set_bounds(2, get_end_mask()) && get_input_tensor(3).has_and_set_bound() &&
+           default_lower_bound_evaluator(this, output_values);
 }
 
 bool op::v1::StridedSlice::evaluate_upper(ov::TensorVector& output_values) const {
-    return strided_slice_input_check(this) && default_upper_bound_evaluator(this, output_values);
+    return indicies_input_has_and_set_bounds(1, get_begin_mask()) &&
+           indicies_input_has_and_set_bounds(2, get_end_mask()) && get_input_tensor(3).has_and_set_bound() &&
+           default_upper_bound_evaluator(this, output_values);
 }
 
 bool op::v1::StridedSlice::evaluate_label(TensorLabelVector& output_labels) const {
-    if (!strided_slice_input_check(this))
-        return false;
-    OPENVINO_SUPPRESS_DEPRECATED_START
-    return default_label_evaluator(this, output_labels);
-    OPENVINO_SUPPRESS_DEPRECATED_END
+    return indicies_input_has_and_set_bounds(1, get_begin_mask()) &&
+           indicies_input_has_and_set_bounds(2, get_end_mask()) && get_input_tensor(3).has_and_set_bound() &&
+           default_label_evaluator(this, {0}, output_labels);
+}
+
+bool op::v1::StridedSlice::constant_fold(OutputVector& output_values, const OutputVector& inputs_values) {
+    auto is_folded = Node::constant_fold(output_values, inputs_values);
+    if (!is_folded) {
+        // If all ignore mask are set for all begin or end then replace this input by dummy constant
+        // to avoid return false from `could_propagate` during bound evaluation (value of const will be ignore).
+        auto get_indicies_input = [&inputs_values](size_t port, AxisSet&& mask) -> Output<Node> {
+            return all_indicies_ignored(inputs_values[port].get_partial_shape(), mask)
+                       ? std::make_shared<op::v0::Constant>(inputs_values[port].get_element_type(),
+                                                            Shape{mask.size()},
+                                                            0)
+                       : inputs_values[port];
+        };
+
+        const auto& begin = get_indicies_input(1, convert_mask_to_axis_set(get_begin_mask()));
+        const auto& end = get_indicies_input(2, convert_mask_to_axis_set(get_end_mask()));
+
+        const auto& output =
+            ((&begin != &inputs_values[1]) || (&end != &inputs_values[2]))
+                ? clone_with_new_inputs(OutputVector{inputs_values[0], begin, end, inputs_values[3]})->output(0)
+                : this->output(0);
+
+        OPENVINO_SUPPRESS_DEPRECATED_START
+        if (const auto c = ov::get_constant_from_source(output)) {
+            OPENVINO_SUPPRESS_DEPRECATED_END
+            output_values[0] = c;
+            is_folded = true;
+        }
+    }
+    return is_folded;
 }
diff --git a/src/core/tests/constant_folding.cpp b/src/core/tests/constant_folding.cpp
index d5ee167dc25..71c1fc52ba9 100644
--- a/src/core/tests/constant_folding.cpp
+++ b/src/core/tests/constant_folding.cpp
@@ -2224,6 +2224,206 @@ TEST(constant_folding, const_strided_slice) {
     ASSERT_EQ(sliced_values, values_out);
 }
 
+TEST(constant_folding, strided_slice_ignored_dynamic_begin_end_values_from_shape_of) {
+    const auto constant =
+        make_shared<op::Constant>(element::i32,
+                                  Shape{1, 1, 2, 4, 2},
+                                  std::vector<int>{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16});
+    constant->set_friendly_name("constant");
+
+    const auto begin_shape = PartialShape{0, -1, 0, 0, 0};
+    const auto p_begin = std::make_shared<op::Parameter>(element::i64, begin_shape);
+    const auto shape_of_begin = std::make_shared<op::ShapeOf>(p_begin);
+    shape_of_begin->set_friendly_name("begin");
+
+    const auto end_shape = PartialShape{-1, 512, 2, 2, 16};
+    const auto p_end = std::make_shared<op::Parameter>(element::i64, end_shape);
+    const auto shape_of_end = std::make_shared<op::ShapeOf>(p_end);
+    shape_of_end->set_friendly_name("end");
+
+    const auto stride = op::Constant::create(element::i64, {5}, {1, 1, 1, 1, 1});
+    stride->set_friendly_name("stride");
+
+    const auto slice = make_shared<op::v1::StridedSlice>(constant,
+                                                         shape_of_begin,
+                                                         shape_of_end,
+                                                         stride,
+                                                         std::vector<int64_t>{0, 1, 0, 0, 0},
+                                                         std::vector<int64_t>{1, 1, 0, 0, 1});
+    slice->set_friendly_name("test");
+
+    auto model = make_shared<ov::Model>(slice, ParameterVector{p_begin, p_end});
+
+    run_constant_folding(model);
+
+    ASSERT_EQ(count_ops_of_type<op::v1::StridedSlice>(model), 0);
+    ASSERT_EQ(count_ops_of_type<op::Constant>(model), 1);
+
+    const auto new_const = get_result_constant(model);
+    ASSERT_TRUE(new_const);
+    check_names(new_const, {"constant", "begin", "end", "stride", "test"});
+    const auto values_out = new_const->get_vector<int>();
+
+    vector<int> sliced_values{1, 2, 3, 4, 9, 10, 11, 12};
+    ASSERT_EQ(sliced_values, values_out);
+}
+
+TEST(constant_folding, strided_slice_all_ignore_mask_set_for_non_parameter_begin_end) {
+    const auto constant =
+        make_shared<op::Constant>(element::i32,
+                                  Shape{1, 1, 2, 4, 2},
+                                  std::vector<int>{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16});
+    constant->set_friendly_name("constant");
+
+    const auto begin_shape = PartialShape{{2, 5}, -1, 10, 1, {0, 200}};
+    const auto p_begin = std::make_shared<op::Parameter>(element::i64, begin_shape);
+    const auto shape_of_begin = std::make_shared<op::ShapeOf>(p_begin);
+    shape_of_begin->set_friendly_name("begin");
+
+    const auto end_shape = PartialShape{-1, 1, {2, 3}, 0, 0};
+    const auto p_end = std::make_shared<op::Parameter>(element::i64, end_shape);
+    const auto shape_of_end = std::make_shared<op::ShapeOf>(p_end);
+    shape_of_end->set_friendly_name("end");
+
+    const auto stride = op::Constant::create(element::i64, {5}, {1, 1, 1, 2, 2});
+    stride->set_friendly_name("stride");
+
+    const auto slice = make_shared<op::v1::StridedSlice>(constant,
+                                                         shape_of_begin,
+                                                         shape_of_end,
+                                                         stride,
+                                                         std::vector<int64_t>{1, 1, 1, 1, 1},
+                                                         std::vector<int64_t>{1, 1, 1, 1, 1});
+    slice->set_friendly_name("test");
+
+    auto model = make_shared<ov::Model>(slice, ParameterVector{p_begin, p_end});
+
+    run_constant_folding(model);
+
+    ASSERT_EQ(count_ops_of_type<op::v1::StridedSlice>(model), 0);
+    ASSERT_EQ(count_ops_of_type<op::Constant>(model), 1);
+
+    const auto new_const = get_result_constant(model);
+    ASSERT_TRUE(new_const);
+    check_names(new_const, {"constant", "begin", "end", "stride", "test"});
+    const auto values_out = new_const->get_vector<int>();
+
+    vector<int> sliced_values{1, 5, 9, 13};
+    ASSERT_EQ(sliced_values, values_out);
+}
+
+TEST(constant_folding, strided_slice_all_ignore_mask_set_for_parameter_begin_end) {
+    const auto constant =
+        make_shared<op::Constant>(element::i32,
+                                  Shape{1, 1, 2, 4, 2},
+                                  std::vector<int>{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16});
+    constant->set_friendly_name("constant");
+
+    const auto begin_shape = PartialShape{{1, 5}};
+    const auto p_begin = std::make_shared<op::Parameter>(element::i64, begin_shape);
+    p_begin->set_friendly_name("begin");
+
+    const auto end_shape = PartialShape{5};
+    const auto p_end = std::make_shared<op::Parameter>(element::i64, end_shape);
+    p_end->set_friendly_name("end");
+
+    const auto stride = op::Constant::create(element::i64, {5}, {1, 1, 1, 2, 2});
+    stride->set_friendly_name("stride");
+
+    const auto slice = make_shared<op::v1::StridedSlice>(constant,
+                                                         p_begin,
+                                                         p_end,
+                                                         stride,
+                                                         std::vector<int64_t>{1, 1, 1, 1, 1},
+                                                         std::vector<int64_t>{1, 1, 1, 1, 1});
+    slice->set_friendly_name("test");
+
+    auto model = make_shared<ov::Model>(slice, ParameterVector{p_begin, p_end});
+
+    run_constant_folding(model);
+
+    ASSERT_EQ(count_ops_of_type<op::v1::StridedSlice>(model), 0);
+    ASSERT_EQ(count_ops_of_type<op::Constant>(model), 1);
+
+    const auto new_const = get_result_constant(model);
+    ASSERT_TRUE(new_const);
+    check_names(new_const, {"constant", "begin", "end", "stride", "test"});
+    const auto values_out = new_const->get_vector<int>();
+
+    vector<int> sliced_values{1, 5, 9, 13};
+    ASSERT_EQ(sliced_values, values_out);
+}
+
+TEST(constant_folding, strided_slice_not_all_ignore_mask_set_for_parameter_begin_end) {
+    const auto constant =
+        make_shared<op::Constant>(element::i32,
+                                  Shape{1, 1, 2, 4, 2},
+                                  std::vector<int>{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16});
+    constant->set_friendly_name("constant");
+
+    const auto begin_shape = PartialShape::dynamic();
+    const auto p_begin = std::make_shared<op::Parameter>(element::i64, begin_shape);
+    p_begin->set_friendly_name("begin");
+
+    const auto end_shape = PartialShape{5};
+    const auto p_end = std::make_shared<op::Parameter>(element::i64, end_shape);
+    p_end->set_friendly_name("end");
+
+    const auto stride = op::Constant::create(element::i64, {5}, {1, 1, 1, 2, 2});
+    stride->set_friendly_name("stride");
+
+    const auto slice = make_shared<op::v1::StridedSlice>(constant,
+                                                         p_begin,
+                                                         p_end,
+                                                         stride,
+                                                         std::vector<int64_t>{1, 1, 1, 1},
+                                                         std::vector<int64_t>{1, 1, 1, 1});
+    slice->set_friendly_name("test");
+
+    auto model = make_shared<ov::Model>(slice, ParameterVector{p_begin, p_end});
+
+    run_constant_folding(model);
+
+    ASSERT_EQ(count_ops_of_type<op::v1::StridedSlice>(model), 1);
+    ASSERT_EQ(count_ops_of_type<op::Constant>(model), 2);
+}
+
+TEST(constant_folding, strided_slice_not_ignored_dynamic_begin_from_shape_of) {
+    const auto constant =
+        make_shared<op::Constant>(element::i32,
+                                  Shape{1, 1, 2, 4, 2},
+                                  std::vector<int>{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16});
+    constant->set_friendly_name("constant");
+
+    const auto begin_shape = PartialShape{0, -1, 0, 0, 0};
+    const auto p_begin = std::make_shared<op::Parameter>(element::i64, begin_shape);
+    const auto shape_of_begin = std::make_shared<op::ShapeOf>(p_begin);
+    shape_of_begin->set_friendly_name("begin");
+
+    const auto end_shape = PartialShape{-1, 512, 2, 2, 16};
+    const auto p_end = std::make_shared<op::Parameter>(element::i64, end_shape);
+    const auto shape_of_end = std::make_shared<op::ShapeOf>(p_end);
+    shape_of_end->set_friendly_name("end");
+
+    const auto stride = op::Constant::create(element::i64, {5}, {1, 1, 1, 1, 1});
+    stride->set_friendly_name("stride");
+
+    const auto slice = make_shared<op::v1::StridedSlice>(constant,
+                                                         shape_of_begin,
+                                                         shape_of_end,
+                                                         stride,
+                                                         std::vector<int64_t>{0, 0, 0, 0, 0},
+                                                         std::vector<int64_t>{1, 1, 0, 0, 1});
+    slice->set_friendly_name("test");
+
+    auto model = make_shared<ov::Model>(slice, ParameterVector{p_begin, p_end});
+
+    run_constant_folding(model);
+
+    ASSERT_EQ(count_ops_of_type<op::v1::StridedSlice>(model), 1);
+    ASSERT_EQ(count_ops_of_type<op::Constant>(model), 2);
+}
+
 TEST(constant_folding, constant_dyn_reshape) {
     Shape shape_in{2, 4};
     vector<float> values_in{0, 1, 2, 3, 4, 5, 6, 7};