diff --git a/src/bindings/python/src/openvino/runtime/opset4/ops.py b/src/bindings/python/src/openvino/runtime/opset4/ops.py
index 1b6ae833079..0ca2a2e27f1 100644
--- a/src/bindings/python/src/openvino/runtime/opset4/ops.py
+++ b/src/bindings/python/src/openvino/runtime/opset4/ops.py
@@ -173,7 +173,7 @@ def acosh(node: NodeInput, name: Optional[str] = None) -> Node:
     :param name: Optional new name for output node.
     :return: New node with arccosh operation applied on it.
     """
-    return _get_node_factory_opset4().create("Acosh", [node])
+    return _get_node_factory_opset4().create("Acosh", as_nodes(node))
 
 
 @nameable_op
@@ -184,7 +184,7 @@ def asinh(node: NodeInput, name: Optional[str] = None) -> Node:
     :param name: Optional new name for output node.
     :return: New node with arcsinh operation applied on it.
     """
-    return _get_node_factory_opset4().create("Asinh", [node])
+    return _get_node_factory_opset4().create("Asinh", as_nodes(node))
 
 
 @nameable_op
@@ -195,7 +195,7 @@ def atanh(node: NodeInput, name: Optional[str] = None) -> Node:
     :param name: Optional new name for output node.
     :return: New node with arctanh operation applied on it.
     """
-    return _get_node_factory_opset4().create("Atanh", [node])
+    return _get_node_factory_opset4().create("Atanh", as_nodes(node))
 
 
 @nameable_op
diff --git a/src/bindings/python/tests/__init__.py b/src/bindings/python/tests/__init__.py
index 95bafaaa833..72d75ee7bd9 100644
--- a/src/bindings/python/tests/__init__.py
+++ b/src/bindings/python/tests/__init__.py
@@ -50,8 +50,6 @@ xfail_issue_90649 = xfail_test(reason="RuntimeError: OV does not support the fol
                                       "MelWeightMatrix, SequenceMap, STFT")
 xfail_issue_35923 = xfail_test(reason="RuntimeError: PReLU without weights is not supported")
 xfail_issue_35927 = xfail_test(reason="RuntimeError: B has zero dimension that is not allowable")
-xfail_issue_36486 = xfail_test(reason="RuntimeError: HardSigmoid operation should be converted "
-                                      "to HardSigmoid_IE")
 xfail_issue_38091 = xfail_test(reason="AssertionError: Mismatched elements")
 xfail_issue_38699 = xfail_test(reason="RuntimeError: OV does not support the following ONNX operations: "
                                       "ai.onnx.preview.training.Gradient")
diff --git a/src/bindings/python/tests/runtime.py b/src/bindings/python/tests/runtime.py
index 9177ed38972..3ca435a5394 100644
--- a/src/bindings/python/tests/runtime.py
+++ b/src/bindings/python/tests/runtime.py
@@ -12,7 +12,7 @@ import numpy as np
 from openvino.runtime import Core
 
 from openvino.runtime.exceptions import UserInputError
-from openvino.runtime import Model, Node, PartialShape, Tensor, Type
+from openvino.runtime import Model, Node, Tensor, Type
 from openvino.runtime.utils.types import NumericData, get_shape, get_dtype
 
 import tests
diff --git a/src/bindings/python/tests/test_graph/test_basic.py b/src/bindings/python/tests/test_graph/test_basic.py
index 43a909242d2..f5a1ef59f78 100644
--- a/src/bindings/python/tests/test_graph/test_basic.py
+++ b/src/bindings/python/tests/test_graph/test_basic.py
@@ -2,7 +2,6 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
-import json
 
 import numpy as np
 import pytest
@@ -15,9 +14,8 @@ from openvino.runtime import Strides, AxisVector, Coordinate, CoordinateDiff
 from openvino.runtime import Tensor, OVAny
 from openvino._pyopenvino import DescriptorTensor
 from openvino.runtime.op import Parameter
-from tests.runtime import get_runtime
-from openvino.runtime.utils.types import get_dtype
-from tests.test_graph.util import run_op_node
+
+from openvino.runtime.utils.types import get_element_type
 
 
 def test_graph_function_api():
@@ -87,76 +85,49 @@ def test_graph_function_api():
         Type.u64,
     ],
 )
-def test_simple_computation_on_ndarrays(dtype):
-    runtime = get_runtime()
-
+def test_simple_model_on_parameters(dtype):
     shape = [2, 2]
     parameter_a = ops.parameter(shape, dtype=dtype, name="A")
     parameter_b = ops.parameter(shape, dtype=dtype, name="B")
     parameter_c = ops.parameter(shape, dtype=dtype, name="C")
     model = (parameter_a + parameter_b) * parameter_c
-    computation = runtime.computation(model, parameter_a, parameter_b, parameter_c)
-
-    np_dtype = get_dtype(dtype) if isinstance(dtype, Type) else dtype
-
-    value_a = np.array([[1, 2], [3, 4]], dtype=np_dtype)
-    value_b = np.array([[5, 6], [7, 8]], dtype=np_dtype)
-    value_c = np.array([[2, 3], [4, 5]], dtype=np_dtype)
-    result = computation(value_a, value_b, value_c)
-    assert np.allclose(result, np.array([[12, 24], [40, 60]], dtype=np_dtype))
-
-    value_a = np.array([[9, 10], [11, 12]], dtype=np_dtype)
-    value_b = np.array([[13, 14], [15, 16]], dtype=np_dtype)
-    value_c = np.array([[5, 4], [3, 2]], dtype=np_dtype)
-    result = computation(value_a, value_b, value_c)
-    assert np.allclose(result, np.array([[110, 96], [78, 56]], dtype=np_dtype))
-
-
-def test_serialization():
-    dtype = np.float32
-    shape = [2, 2]
-    parameter_a = ops.parameter(shape, dtype=dtype, name="A")
-    parameter_b = ops.parameter(shape, dtype=dtype, name="B")
-    parameter_c = ops.parameter(shape, dtype=dtype, name="C")
-    model = (parameter_a + parameter_b) * parameter_c
-
-    runtime = get_runtime()
-    computation = runtime.computation(model, parameter_a, parameter_b, parameter_c)
-    try:
-        serialized = computation.serialize(2)
-        serial_json = json.loads(serialized)
-
-        assert serial_json[0]["name"] != ""
-        assert 10 == len(serial_json[0]["ops"])
-    except Exception:
-        pass
+    expected_type = dtype if isinstance(dtype, Type) else get_element_type(dtype)
+    assert model.get_type_name() == "Multiply"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == expected_type
+    assert list(model.get_output_shape(0)) == [2, 2]
 
 
 def test_broadcast_1():
-    input_data = np.array([1, 2, 3], dtype=np.int32)
-
+    input_data = ops.parameter((3,), name="input_data", dtype=np.int32)
     new_shape = [3, 3]
-    expected = [[1, 2, 3], [1, 2, 3], [1, 2, 3]]
-    result = run_op_node([input_data], ops.broadcast, new_shape)
-    assert np.allclose(result, expected)
+    node = ops.broadcast(input_data, new_shape)
+    assert node.get_type_name() == "Broadcast"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == Type.i32
+    assert list(node.get_output_shape(0)) == [3, 3]
 
 
 def test_broadcast_2():
-    input_data = np.arange(4, dtype=np.int32)
+    input_data = ops.parameter((4,), name="input_data", dtype=np.int32)
     new_shape = [3, 4, 2, 4]
-    expected = np.broadcast_to(input_data, new_shape)
-    result = run_op_node([input_data], ops.broadcast, new_shape)
-    assert np.allclose(result, expected)
+    expected_shape = np.broadcast_to(input_data, new_shape).shape
+    node = ops.broadcast(input_data, new_shape)
+    assert node.get_type_name() == "Broadcast"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == Type.i32
+    assert list(node.get_output_shape(0)) == list(expected_shape)
 
 
 def test_broadcast_3():
-    input_data = np.array([1, 2, 3], dtype=np.int32)
+    input_data = ops.parameter((3,), name="input_data", dtype=np.int32)
     new_shape = [3, 3]
     axis_mapping = [0]
-    expected = [[1, 1, 1], [2, 2, 2], [3, 3, 3]]
-
-    result = run_op_node([input_data], ops.broadcast, new_shape, axis_mapping, "EXPLICIT")
-    assert np.allclose(result, expected)
+    node = ops.broadcast(input_data, new_shape, axis_mapping, "EXPLICIT")
+    assert node.get_type_name() == "Broadcast"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == Type.i32
+    assert list(node.get_output_shape(0)) == [3, 3]
 
 
 @pytest.mark.parametrize(
@@ -164,10 +135,11 @@ def test_broadcast_3():
     [(bool, np.zeros((2, 2), dtype=np.int32)), ("boolean", np.zeros((2, 2), dtype=np.int32))],
 )
 def test_convert_to_bool(destination_type, input_data):
-    expected = np.array(input_data, dtype=bool)
-    result = run_op_node([input_data], ops.convert, destination_type)
-    assert np.allclose(result, expected)
-    assert np.array(result).dtype == bool
+    node = ops.convert(input_data, destination_type)
+    assert node.get_type_name() == "Convert"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == Type.boolean
+    assert list(node.get_output_shape(0)) == [2, 2]
 
 
 @pytest.mark.parametrize(
@@ -182,10 +154,11 @@ def test_convert_to_bool(destination_type, input_data):
 def test_convert_to_float(destination_type, rand_range, in_dtype, expected_type):
     np.random.seed(133391)
     input_data = np.random.randint(*rand_range, size=(2, 2), dtype=in_dtype)
-    expected = np.array(input_data, dtype=expected_type)
-    result = run_op_node([input_data], ops.convert, destination_type)
-    assert np.allclose(result, expected)
-    assert np.array(result).dtype == expected_type
+    node = ops.convert(input_data, destination_type)
+    assert node.get_type_name() == "Convert"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(expected_type)
+    assert list(node.get_output_shape(0)) == [2, 2]
 
 
 @pytest.mark.parametrize(
@@ -205,10 +178,11 @@ def test_convert_to_int(destination_type, expected_type):
     np.random.seed(133391)
     random_data = np.random.rand(2, 3, 4) * 16
     input_data = (np.ceil(-8 + random_data)).astype(expected_type)
-    expected = np.array(input_data, dtype=expected_type)
-    result = run_op_node([input_data], ops.convert, destination_type)
-    assert np.allclose(result, expected)
-    assert np.array(result).dtype == expected_type
+    node = ops.convert(input_data, destination_type)
+    assert node.get_type_name() == "Convert"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(expected_type)
+    assert list(node.get_output_shape(0)) == [2, 3, 4]
 
 
 @pytest.mark.parametrize(
@@ -227,23 +201,11 @@ def test_convert_to_int(destination_type, expected_type):
 def test_convert_to_uint(destination_type, expected_type):
     np.random.seed(133391)
     input_data = np.ceil(np.random.rand(2, 3, 4) * 16).astype(expected_type)
-    expected = np.array(input_data, dtype=expected_type)
-    result = run_op_node([input_data], ops.convert, destination_type)
-    assert np.allclose(result, expected)
-    assert np.array(result).dtype == expected_type
-
-
-def test_bad_data_shape():
-    param_a = ops.parameter(shape=[2, 2], name="A", dtype=np.float32)
-    param_b = ops.parameter(shape=[2, 2], name="B")
-    model = param_a + param_b
-    runtime = get_runtime()
-    computation = runtime.computation(model, param_a, param_b)
-
-    value_a = np.array([[1, 2]], dtype=np.float32)
-    value_b = np.array([[5, 6], [7, 8]], dtype=np.float32)
-    with pytest.raises(RuntimeError):
-        computation(value_a, value_b)
+    node = ops.convert(input_data, destination_type)
+    assert node.get_type_name() == "Convert"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(expected_type)
+    assert list(node.get_output_shape(0)) == [2, 3, 4]
 
 
 def test_constant_get_data_bool():
@@ -289,41 +251,41 @@ def test_constant_get_data_unsigned_integer(data_type):
 
 
 def test_set_argument():
-    runtime = get_runtime()
-
     data1 = np.array([1, 2, 3])
     data2 = np.array([4, 5, 6])
     data3 = np.array([7, 8, 9])
 
     node1 = ops.constant(data1, dtype=np.float32)
     node2 = ops.constant(data2, dtype=np.float32)
-    node3 = ops.constant(data3, dtype=np.float32)
+    node3 = ops.constant(data3, dtype=np.float64)
+    node4 = ops.constant(data3, dtype=np.float64)
     node_add = ops.add(node1, node2)
 
     # Original arguments
-    computation = runtime.computation(node_add)
-    output = computation()
-    assert np.allclose(data1 + data2, output)
-
-    # Arguments changed by set_argument
-    node_add.set_argument(1, node3.output(0))
-    output = computation()
-    assert np.allclose(data1 + data3, output)
+    node_inputs = node_add.inputs()
+    assert node_inputs[0].get_element_type() == Type.f32
+    assert node_inputs[1].get_element_type() == Type.f32
+    assert len(node_inputs) == 2
 
     # Arguments changed by set_argument
     node_add.set_argument(0, node3.output(0))
-    output = computation()
-    assert np.allclose(data3 + data3, output)
+    node_add.set_argument(1, node4.output(0))
+    node_inputs = node_add.inputs()
+    assert node_inputs[0].get_element_type() == Type.f64
+    assert node_inputs[1].get_element_type() == Type.f64
+    assert len(node_inputs) == 2
 
     # Arguments changed by set_argument(OutputVector)
-    node_add.set_arguments([node2.output(0), node3.output(0)])
-    output = computation()
-    assert np.allclose(data2 + data3, output)
+    node_add.set_arguments([node1.output(0), node2.output(0)])
+    assert node_inputs[0].get_element_type() == Type.f32
+    assert node_inputs[1].get_element_type() == Type.f32
+    assert len(node_inputs) == 2
 
     # Arguments changed by set_arguments(NodeVector)
-    node_add.set_arguments([node1, node2])
-    output = computation()
-    assert np.allclose(data1 + data2, output)
+    node_add.set_arguments([node3, node4])
+    assert node_inputs[0].get_element_type() == Type.f64
+    assert node_inputs[1].get_element_type() == Type.f64
+    assert len(node_inputs) == 2
 
 
 def test_clone_model():
@@ -351,9 +313,12 @@ def test_clone_model():
 
 
 def test_result():
-    node = np.array([[11, 10], [1, 8], [3, 4]], dtype=np.float32)
-    result = run_op_node([node], ops.result)
-    assert np.allclose(result, node)
+    input_data = np.array([[11, 10], [1, 8], [3, 4]], dtype=np.float32)
+    node = ops.result(input_data)
+    assert node.get_type_name() == "Result"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == Type.f32
+    assert list(node.get_output_shape(0)) == [3, 2]
 
 
 def test_node_friendly_name():
@@ -566,11 +531,10 @@ def test_multiple_outputs():
     split_first_output = split.output(0)
     relu = ops.relu(split_first_output)
 
-    runtime = get_runtime()
-    computation = runtime.computation(relu, test_param)
-    output = computation(input_data)
-
-    assert np.equal(output, expected_output).all()
+    assert relu.get_type_name() == "Relu"
+    assert relu.get_output_size() == 1
+    assert relu.get_output_element_type(0) == Type.f32
+    assert list(relu.get_output_shape(0)) == [4, 2]
 
 
 def test_sink_function_ctor():
diff --git a/src/bindings/python/tests/test_graph/test_convolution.py b/src/bindings/python/tests/test_graph/test_convolution.py
index 2d2856339b4..1a4533b4623 100644
--- a/src/bindings/python/tests/test_graph/test_convolution.py
+++ b/src/bindings/python/tests/test_graph/test_convolution.py
@@ -3,11 +3,19 @@
 # SPDX-License-Identifier: Apache-2.0
 
 import numpy as np
+import pytest
 
+from openvino.runtime import Type
 import openvino.runtime.opset8 as ov
 
 
-def test_convolution_2d():
+@pytest.mark.parametrize(("strides", "pads_begin", "pads_end", "dilations", "expected_shape"), [
+    (np.array([1, 1]), np.array([1, 1]), np.array([1, 1]), np.array([1, 1]), [1, 1, 9, 9]),
+    (np.array([1, 1]), np.array([0, 0]), np.array([0, 0]), np.array([1, 1]), [1, 1, 7, 7]),
+    (np.array([2, 2]), np.array([0, 0]), np.array([0, 0]), np.array([1, 1]), [1, 1, 4, 4]),
+    (np.array([1, 1]), np.array([0, 0]), np.array([0, 0]), np.array([2, 2]), [1, 1, 5, 5]),
+])
+def test_convolution_2d(strides, pads_begin, pads_end, dilations, expected_shape):
 
     # input_x should have shape N(batch) x C x H x W
     input_x = ov.parameter((1, 1, 9, 9), name="input_x", dtype=np.float32)
@@ -15,53 +23,11 @@ def test_convolution_2d():
     # filter weights should have shape M x C x kH x kW
     input_filter = ov.parameter((1, 1, 3, 3), name="input_filter", dtype=np.float32)
 
-    strides = np.array([1, 1])
-    pads_begin = np.array([1, 1])
-    pads_end = np.array([1, 1])
-    dilations = np.array([1, 1])
-    expected_shape = [1, 1, 9, 9]
-
-    # convolution with padding=1 should produce 9 x 9 output:
-    node = ov.convolution(input_x, input_filter, strides, pads_begin, pads_end, dilations)
-    assert node.get_type_name() == "Convolution"
-    assert node.get_output_size() == 1
-    assert list(node.get_output_shape(0)) == expected_shape
-
-    # convolution with padding=0 should produce 7 x 7 output:
-    strides = np.array([1, 1])
-    pads_begin = np.array([0, 0])
-    pads_end = np.array([0, 0])
-    dilations = np.array([1, 1])
-    expected_shape = [1, 1, 7, 7]
-
-    node = ov.convolution(input_x, input_filter, strides, pads_begin, pads_end, dilations)
-    assert node.get_type_name() == "Convolution"
-    assert node.get_output_size() == 1
-    assert list(node.get_output_shape(0)) == expected_shape
-
-    strides = np.array([2, 2])
-    pads_begin = np.array([0, 0])
-    pads_end = np.array([0, 0])
-    dilations = np.array([1, 1])
-    expected_shape = [1, 1, 4, 4]
-
-    # convolution with strides=2 should produce 4 x 4 output:
-    node = ov.convolution(input_x, input_filter, strides, pads_begin, pads_end, dilations)
-    assert node.get_type_name() == "Convolution"
-    assert node.get_output_size() == 1
-    assert list(node.get_output_shape(0)) == expected_shape
-
-    strides = np.array([1, 1])
-    pads_begin = np.array([0, 0])
-    pads_end = np.array([0, 0])
-    dilations = np.array([2, 2])
-    expected_shape = [1, 1, 5, 5]
-
-    # convolution with dilation=2 should produce 5 x 5 output:
     node = ov.convolution(input_x, input_filter, strides, pads_begin, pads_end, dilations)
     assert node.get_type_name() == "Convolution"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_convolution_backprop_data():
@@ -80,6 +46,7 @@ def test_convolution_backprop_data():
     assert deconvolution.get_type_name() == "ConvolutionBackpropData"
     assert deconvolution.get_output_size() == 1
     assert list(deconvolution.get_output_shape(0)) == expected_shape
+    assert deconvolution.get_output_element_type(0) == Type.f32
 
 
 def test_convolution_v1():
@@ -95,3 +62,4 @@ def test_convolution_v1():
     assert node.get_type_name() == "Convolution"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_create_op.py b/src/bindings/python/tests/test_graph/test_create_op.py
index 0ddbf73acb0..5def6f7824f 100644
--- a/src/bindings/python/tests/test_graph/test_create_op.py
+++ b/src/bindings/python/tests/test_graph/test_create_op.py
@@ -2203,7 +2203,7 @@ def test_interpolate_opset10(dtype, expected_shape, shape_calculation_mode):
 
 def test_is_finite_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ov.parameter(input_shape, float, name="InputData")
+    input_node = ov.parameter(input_shape, np.float32, name="InputData")
     node = ov_opset10.is_finite(input_node)
 
     assert node.get_type_name() == "IsFinite"
@@ -2214,7 +2214,7 @@ def test_is_finite_opset10():
 
 def test_is_inf_opset10_default():
     input_shape = [2, 2, 2, 2]
-    input_node = ov.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ov.parameter(input_shape, dtype=np.float32, name="InputData")
     node = ov_opset10.is_inf(input_node)
 
     assert node.get_type_name() == "IsInf"
@@ -2228,7 +2228,7 @@ def test_is_inf_opset10_default():
 
 def test_is_inf_opset10_custom_attribute():
     input_shape = [2, 2, 2]
-    input_node = ov.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ov.parameter(input_shape, dtype=np.float32, name="InputData")
     attributes = {
         "detect_positive": False,
     }
@@ -2245,7 +2245,7 @@ def test_is_inf_opset10_custom_attribute():
 
 def test_is_inf_opset10_custom_all_attributes():
     input_shape = [2, 2, 2]
-    input_node = ov.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ov.parameter(input_shape, dtype=np.float32, name="InputData")
     attributes = {
         "detect_negative": False,
         "detect_positive": True,
@@ -2263,7 +2263,7 @@ def test_is_inf_opset10_custom_all_attributes():
 
 def test_is_nan_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ov.parameter(input_shape, float, name="InputData")
+    input_node = ov.parameter(input_shape, np.float32, name="InputData")
     node = ov_opset10.is_nan(input_node)
 
     assert node.get_type_name() == "IsNaN"
@@ -2274,7 +2274,7 @@ def test_is_nan_opset10():
 
 def test_unique_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ov.parameter(input_shape, float, name="input_data")
+    input_node = ov.parameter(input_shape, np.float32, name="input_data")
     axis = ov.constant([1], np.int32, [1])
 
     node = ov_opset10.unique(input_node, axis, False, "i32")
diff --git a/src/bindings/python/tests/test_graph/test_data_movement.py b/src/bindings/python/tests/test_graph/test_data_movement.py
index 260de21ffbc..7b7ecf672e7 100644
--- a/src/bindings/python/tests/test_graph/test_data_movement.py
+++ b/src/bindings/python/tests/test_graph/test_data_movement.py
@@ -22,6 +22,7 @@ def test_reverse_sequence():
     assert model.get_type_name() == "ReverseSequence"
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == expected_shape
+    assert model.get_output_element_type(0) == Type.i32
 
 
 def test_pad_edge():
@@ -35,6 +36,7 @@ def test_pad_edge():
     assert model.get_type_name() == "Pad"
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == expected_shape
+    assert model.get_output_element_type(0) == Type.i32
 
 
 def test_pad_constant():
@@ -48,6 +50,7 @@ def test_pad_constant():
     assert model.get_type_name() == "Pad"
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == expected_shape
+    assert model.get_output_element_type(0) == Type.i32
 
 
 def test_select():
@@ -60,6 +63,7 @@ def test_select():
     assert node.get_type_name() == "Select"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.i32
 
 
 def test_gather_v8_nd():
diff --git a/src/bindings/python/tests/test_graph/test_dft.py b/src/bindings/python/tests/test_graph/test_dft.py
index 06a1a3357b9..f5522f2df28 100644
--- a/src/bindings/python/tests/test_graph/test_dft.py
+++ b/src/bindings/python/tests/test_graph/test_dft.py
@@ -2,6 +2,7 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
+from openvino.runtime import Type
 import openvino.runtime.opset9 as ov
 import numpy as np
 
@@ -23,6 +24,7 @@ def test_dft_1d():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_2d():
@@ -37,6 +39,7 @@ def test_dft_2d():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_3d():
@@ -51,6 +54,7 @@ def test_dft_3d():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_1d_signal_size():
@@ -66,6 +70,7 @@ def test_dft_1d_signal_size():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_2d_signal_size_1():
@@ -81,6 +86,7 @@ def test_dft_2d_signal_size_1():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_2d_signal_size_2():
@@ -96,6 +102,7 @@ def test_dft_2d_signal_size_2():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_dft_3d_signal_size():
@@ -111,3 +118,4 @@ def test_dft_3d_signal_size():
     assert dft_node.get_type_name() == "DFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == expected_shape
+    assert dft_node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_gather.py b/src/bindings/python/tests/test_graph/test_gather.py
index 108f631d7a0..c9b76fc578d 100644
--- a/src/bindings/python/tests/test_graph/test_gather.py
+++ b/src/bindings/python/tests/test_graph/test_gather.py
@@ -2,6 +2,7 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
+from openvino.runtime import Type
 import openvino.runtime.opset8 as ov
 import numpy as np
 
@@ -16,6 +17,7 @@ def test_gather():
     assert node.get_type_name() == "Gather"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_gather_with_scalar_axis():
@@ -28,6 +30,7 @@ def test_gather_with_scalar_axis():
     assert node.get_type_name() == "Gather"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_gather_batch_dims_1():
@@ -41,6 +44,7 @@ def test_gather_batch_dims_1():
     assert node.get_type_name() == "Gather"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_gather_negative_indices():
@@ -53,6 +57,7 @@ def test_gather_negative_indices():
     assert node.get_type_name() == "Gather"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_gather_batch_dims_1_negative_indices():
@@ -66,3 +71,4 @@ def test_gather_batch_dims_1_negative_indices():
     assert node.get_type_name() == "Gather"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_idft.py b/src/bindings/python/tests/test_graph/test_idft.py
index fda35a17f69..dfeb77df598 100644
--- a/src/bindings/python/tests/test_graph/test_idft.py
+++ b/src/bindings/python/tests/test_graph/test_idft.py
@@ -2,6 +2,7 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
+from openvino.runtime import Type
 import openvino.runtime.opset8 as ov
 import numpy as np
 
@@ -23,6 +24,7 @@ def test_idft_1d():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_2d():
@@ -37,6 +39,7 @@ def test_idft_2d():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_3d():
@@ -51,6 +54,7 @@ def test_idft_3d():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_1d_signal_size():
@@ -66,6 +70,7 @@ def test_idft_1d_signal_size():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_2d_signal_size_1():
@@ -81,6 +86,7 @@ def test_idft_2d_signal_size_1():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_2d_signal_size_2():
@@ -96,6 +102,7 @@ def test_idft_2d_signal_size_2():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
 
 
 def test_idft_3d_signal_size():
@@ -111,3 +118,4 @@ def test_idft_3d_signal_size():
     assert dft_node.get_type_name() == "IDFT"
     assert dft_node.get_output_size() == 1
     assert list(dft_node.get_output_shape(0)) == list(expected_results.shape)
+    assert dft_node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_if.py b/src/bindings/python/tests/test_graph/test_if.py
index 2db365c15c9..def8739a2d7 100644
--- a/src/bindings/python/tests/test_graph/test_if.py
+++ b/src/bindings/python/tests/test_graph/test_if.py
@@ -6,8 +6,6 @@ import numpy as np
 import openvino.runtime.opset8 as ov
 from openvino.runtime import Model
 
-from tests.runtime import get_runtime
-
 from openvino.runtime.op.util import InvariantInputDescription, BodyOutputDescription
 
 
@@ -149,34 +147,33 @@ def check_results(results, expected_results):
 
 def check_if(if_model, cond_val, exp_results):
     last_node = if_model(cond_val)
-    runtime = get_runtime()
-    computation = runtime.computation(last_node)
-    results = computation()
-    check_results(results, exp_results)
+    assert last_node.get_type_name() == exp_results[0]
+    assert last_node.get_output_size() == exp_results[1]
+    assert list(last_node.get_output_shape(0)) == exp_results[2]
 
 
 def test_if_with_two_outputs():
     check_if(create_simple_if_with_two_outputs, True,
-             [np.array([10], dtype=np.float32), np.array([-20], dtype=np.float32)])
+             ["If", 2, []])
     check_if(create_simple_if_with_two_outputs, False,
-             [np.array([17], dtype=np.float32), np.array([16], dtype=np.float32)])
+             ["If", 2, []])
 
 
 def test_diff_if_with_two_outputs():
     check_if(create_diff_if_with_two_outputs, True,
-             [np.array([10], dtype=np.float32), np.array([6, 4], dtype=np.float32)])
+             ["If", 2, []])
     check_if(create_diff_if_with_two_outputs, False,
-             [np.array([4], dtype=np.float32), np.array([12, 16], dtype=np.float32)])
+             ["If", 2, []])
 
 
 def test_simple_if():
-    check_if(simple_if, True, [np.array([6, 4], dtype=np.float32)])
-    check_if(simple_if, False, [np.array([5, 5], dtype=np.float32)])
+    check_if(simple_if, True, ["Relu", 1, [2]])
+    check_if(simple_if, False, ["Relu", 1, [2]])
 
 
 def test_simple_if_without_body_parameters():
-    check_if(simple_if_without_parameters, True, [np.array([0.7], dtype=np.float32)])
-    check_if(simple_if_without_parameters, False, [np.array([9.0], dtype=np.float32)])
+    check_if(simple_if_without_parameters, True, ["Relu", 1, []])
+    check_if(simple_if_without_parameters, False, ["Relu", 1, []])
 
 
 def test_simple_if_basic():
diff --git a/src/bindings/python/tests/test_graph/test_normalization.py b/src/bindings/python/tests/test_graph/test_normalization.py
index e0ef81fc21f..bbf48728de5 100644
--- a/src/bindings/python/tests/test_graph/test_normalization.py
+++ b/src/bindings/python/tests/test_graph/test_normalization.py
@@ -4,6 +4,7 @@
 
 import numpy as np
 
+from openvino.runtime import Type
 import openvino.runtime.opset8 as ov
 
 
@@ -15,12 +16,14 @@ def test_lrn():
     assert model.get_type_name() == "LRN"
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == [2, 3, 2, 1]
+    assert model.get_output_element_type(0) == Type.f32
 
     # Test LRN default parameter values
     model = ov.lrn(ov.constant(input_image), ov.constant(axes))
     assert model.get_type_name() == "LRN"
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == [2, 3, 2, 1]
+    assert model.get_output_element_type(0) == Type.f32
 
 
 def test_lrn_factory():
@@ -36,6 +39,7 @@ def test_lrn_factory():
     assert node.get_type_name() == "LRN"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_batch_norm():
@@ -51,6 +55,7 @@ def test_batch_norm():
     assert node.get_type_name() == "BatchNormInference"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_mvn_no_variance():
@@ -66,6 +71,7 @@ def test_mvn_no_variance():
     assert node.get_type_name() == "MVN"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_mvn():
@@ -81,3 +87,4 @@ def test_mvn():
     assert node.get_type_name() == "MVN"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == expected_shape
+    assert node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_ops.py b/src/bindings/python/tests/test_graph/test_ops.py
index be0fa20e142..f13ab706da0 100644
--- a/src/bindings/python/tests/test_graph/test_ops.py
+++ b/src/bindings/python/tests/test_graph/test_ops.py
@@ -6,9 +6,8 @@
 import numpy as np
 
 import openvino.runtime.opset8 as ov
-from openvino.runtime import AxisSet, Model, Shape, Type
+from openvino.runtime import AxisSet, Shape, Type
 from openvino.runtime.op import Constant, Parameter
-from tests.runtime import get_runtime
 
 
 def binary_op(op_str, a, b):
@@ -81,47 +80,40 @@ def binary_op_ref(op_str, a, b):
         return np.power(a, b)
 
 
-def binary_op_exec(op_str):
+def binary_op_exec(op_str, expected_ov_str=None):
+    if not expected_ov_str:
+        expected_ov_str = op_str
 
     element_type = Type.f32
     shape = Shape([2, 2])
     A = Parameter(element_type, shape)
     B = Parameter(element_type, shape)
-    parameter_list = [A, B]
-    function = Model([binary_op(op_str, A, B)], parameter_list, "test")
+    node = binary_op(op_str, A, B)
 
-    a_arr = np.array([[1, 6], [7, 4]], dtype=np.float32)
-    b_arr = np.array([[5, 2], [3, 8]], dtype=np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, A, B)
-    result = computation(a_arr, b_arr)[0]
-
-    expected = binary_op_ref(op_str, a_arr, b_arr)
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == expected_ov_str
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [2, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
-def binary_op_comparison(op_str):
+def binary_op_comparison(op_str, expected_ov_str=None):
+    if not expected_ov_str:
+        expected_ov_str = op_str
 
     element_type = Type.f32
     shape = Shape([2, 2])
     A = Parameter(element_type, shape)
     B = Parameter(element_type, shape)
-    parameter_list = [A, B]
-    function = Model([binary_op(op_str, A, B)], parameter_list, "test")
-    a_arr = np.array([[1, 5], [3, 2]], dtype=np.float32)
-    b_arr = np.array([[2, 4], [3, 1]], dtype=np.float32)
+    node = binary_op(op_str, A, B)
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, A, B)
-    result = computation(a_arr, b_arr)[0]
-
-    expected = binary_op_ref(op_str, a_arr, b_arr)
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == expected_ov_str
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [2, 2]
+    assert node.get_output_element_type(0) == Type.boolean
 
 
 def test_add():
-    binary_op_exec("+")
+    binary_op_exec("+", "Add")
 
 
 def test_add_op():
@@ -129,27 +121,27 @@ def test_add_op():
 
 
 def test_sub():
-    binary_op_exec("-")
+    binary_op_exec("-", "Subtract")
 
 
 def test_sub_op():
-    binary_op_exec("Sub")
+    binary_op_exec("Sub", "Subtract")
 
 
 def test_mul():
-    binary_op_exec("*")
+    binary_op_exec("*", "Multiply")
 
 
 def test_mul_op():
-    binary_op_exec("Mul")
+    binary_op_exec("Mul", "Multiply")
 
 
 def test_div():
-    binary_op_exec("/")
+    binary_op_exec("/", "Divide")
 
 
 def test_div_op():
-    binary_op_exec("Div")
+    binary_op_exec("Div", "Divide")
 
 
 def test_maximum():
@@ -169,7 +161,7 @@ def test_greater():
 
 
 def test_greater_eq():
-    binary_op_comparison("GreaterEq")
+    binary_op_comparison("GreaterEq", "GreaterEqual")
 
 
 def test_less():
@@ -177,7 +169,7 @@ def test_less():
 
 
 def test_less_eq():
-    binary_op_comparison("LessEq")
+    binary_op_comparison("LessEq", "LessEqual")
 
 
 def test_not_equal():
@@ -191,23 +183,12 @@ def test_add_with_mul():
     A = Parameter(element_type, shape)
     B = Parameter(element_type, shape)
     C = Parameter(element_type, shape)
-    parameter_list = [A, B, C]
-    function = Model([ov.multiply(ov.add(A, B), C)], parameter_list, "test")
+    node = ov.multiply(ov.add(A, B), C)
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, A, B, C)
-    result = computation(
-        np.array([1, 2, 3, 4], dtype=np.float32),
-        np.array([5, 6, 7, 8], dtype=np.float32),
-        np.array([9, 10, 11, 12], dtype=np.float32),
-    )[0]
-
-    a_arr = np.array([1, 2, 3, 4], dtype=np.float32)
-    b_arr = np.array([5, 6, 7, 8], dtype=np.float32)
-    c_arr = np.array([9, 10, 11, 12], dtype=np.float32)
-    result_arr_ref = (a_arr + b_arr) * c_arr
-
-    assert np.allclose(result, result_arr_ref)
+    assert node.get_type_name() == "Multiply"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [4]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def unary_op(op_str, a):
@@ -298,22 +279,22 @@ def unary_op_ref(op_str, a):
         return np.tanh(a)
 
 
-def unary_op_exec(op_str, input_list):
+def unary_op_exec(op_str, input_list, expected_ov_str=None):
     """
     input_list needs to have deep length of 4
     """
+    if not expected_ov_str:
+        expected_ov_str = op_str
+
     element_type = Type.f32
     shape = Shape(np.array(input_list).shape)
     A = Parameter(element_type, shape)
-    parameter_list = [A]
-    function = Model([unary_op(op_str, A)], parameter_list, "test")
+    node = unary_op(op_str, A)
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(np.array(input_list, dtype=np.float32))[0]
-
-    expected = unary_op_ref(op_str, np.array(input_list, dtype=np.float32))
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == expected_ov_str
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == list(shape)
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_abs():
@@ -385,19 +366,19 @@ def test_floor():
 def test_log():
     input_list = [1, 2, 3, 4]
     op_str = "log"
-    unary_op_exec(op_str, input_list)
+    unary_op_exec(op_str, input_list, "Log")
 
 
 def test_exp():
     input_list = [-1, 0, 1, 2]
     op_str = "exp"
-    unary_op_exec(op_str, input_list)
+    unary_op_exec(op_str, input_list, "Exp")
 
 
 def test_negative():
     input_list = [-1, 0, 1, 2]
     op_str = "negative"
-    unary_op_exec(op_str, input_list)
+    unary_op_exec(op_str, input_list, "Negative")
 
 
 def test_sign():
@@ -437,95 +418,62 @@ def test_tanh():
 
 
 def test_reshape():
-
     element_type = Type.f32
     shape = Shape([2, 3])
     A = Parameter(element_type, shape)
-    parameter_list = [A]
-    function = Model([ov.reshape(A, Shape([3, 2]), special_zero=False)], parameter_list, "test")
+    node = ov.reshape(A, Shape([3, 2]), special_zero=False)
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(np.array(np.array([[1, 2, 3], [4, 5, 6]], dtype=np.float32), dtype=np.float32))[0]
-
-    expected = np.reshape(np.array([[1, 2, 3], [4, 5, 6]], dtype=np.float32), (3, 2))
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == "Reshape"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 2]
+    assert node.get_output_element_type(0) == element_type
 
 
 def test_broadcast():
-
     element_type = Type.f32
     A = Parameter(element_type, Shape([3]))
-    parameter_list = [A]
-    function = Model([ov.broadcast(A, [3, 3])], parameter_list, "test")
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(np.array([1, 2, 3], dtype=np.float32))[0]
-
-    a_arr = np.array([[0], [0], [0]], dtype=np.float32)
-    b_arr = np.array([[1, 2, 3]], dtype=np.float32)
-    expected = np.add(a_arr, b_arr)
-    assert np.allclose(result, expected)
+    node = ov.broadcast(A, [3, 3])
+    assert node.get_type_name() == "Broadcast"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 3]
+    assert node.get_output_element_type(0) == element_type
 
 
 def test_constant():
     element_type = Type.f32
-    parameter_list = []
-    function = Model([Constant(element_type, Shape([3, 3]), list(range(9)))], parameter_list, "test")
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation()[0]
-
-    expected = np.arange(9).reshape(3, 3)
-    assert np.allclose(result, expected)
+    node = Constant(element_type, Shape([3, 3]), list(range(9)))
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 3]
+    assert node.get_output_element_type(0) == element_type
 
 
 def test_constant_opset_ov_type():
-    parameter_list = []
-    function = Model([ov.constant(np.arange(9).reshape(3, 3), Type.f32)], parameter_list, "test")
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation()[0]
-
-    expected = np.arange(9).reshape(3, 3)
-    assert np.allclose(result, expected)
+    node = ov.constant(np.arange(9).reshape(3, 3), Type.f32)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 3]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_constant_opset_numpy_type():
-    parameter_list = []
-    function = Model([ov.constant(np.arange(9).reshape(3, 3), np.float32)], parameter_list, "test")
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation()[0]
-
-    expected = np.arange(9).reshape(3, 3)
-    assert np.allclose(result, expected)
+    node = ov.constant(np.arange(9).reshape(3, 3), np.float32)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 3]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_concat():
-
     element_type = Type.f32
     A = Parameter(element_type, Shape([1, 2]))
     B = Parameter(element_type, Shape([1, 2]))
     C = Parameter(element_type, Shape([1, 2]))
-    parameter_list = [A, B, C]
-    axis = 0
-    function = Model([ov.concat([A, B, C], axis)], parameter_list, "test")
-
-    a_arr = np.array([[1, 2]], dtype=np.float32)
-    b_arr = np.array([[5, 6]], dtype=np.float32)
-    c_arr = np.array([[7, 8]], dtype=np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(a_arr, b_arr, c_arr)[0]
-
-    expected = np.concatenate((a_arr, b_arr, c_arr), axis)
-    assert np.allclose(result, expected)
+    node = ov.concat([A, B, C], axis=0)
+    assert node.get_type_name() == "Concat"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [3, 2]
+    assert node.get_output_element_type(0) == element_type
 
 
 def test_axisset():
@@ -549,29 +497,17 @@ def test_select():
     A = Parameter(Type.boolean, Shape([1, 2]))
     B = Parameter(element_type, Shape([1, 2]))
     C = Parameter(element_type, Shape([1, 2]))
-    parameter_list = [A, B, C]
+    node = ov.select(A, B, C)
+    assert node.get_type_name() == "Select"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 2]
+    assert node.get_output_element_type(0) == element_type
 
-    function = Model([ov.select(A, B, C)], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(
-        np.array([[True, False]], dtype=bool),
-        np.array([[5, 6]], dtype=np.float32),
-        np.array([[7, 8]], dtype=np.float32),
-    )[0]
-
-    expected = np.array([[5, 8]])
-    assert np.allclose(result, expected)
-
-def test_max_pool():
-    # test 1d
+def test_max_pool_1d():
     element_type = Type.f32
     shape = Shape([1, 1, 10])
     A = Parameter(element_type, shape)
-    parameter_list = [A]
-
-    input_arr = np.arange(10, dtype=np.float32).reshape([1, 1, 10])
     window_shape = [3]
 
     strides = [1] * len(window_shape)
@@ -593,19 +529,25 @@ def test_max_pool():
         auto_pad,
         idx_elem_type,
     )
-    function = Model([model], parameter_list, "test")
+    assert model.get_type_name() == "MaxPool"
+    assert model.get_output_size() == 2
+    assert list(model.get_output_shape(0)) == [1, 1, 8]
+    assert list(model.get_output_shape(1)) == [1, 1, 8]
+    assert model.get_output_element_type(0) == element_type
+    assert model.get_output_element_type(1) == Type.i32
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(input_arr)[0]
-
-    expected = (np.arange(8) + 2).reshape(1, 1, 8)
-    assert np.allclose(result, expected)
-
-    # test 1d with strides
+def test_max_pool_1d_with_strides():
+    element_type = Type.f32
+    shape = Shape([1, 1, 10])
+    A = Parameter(element_type, shape)
+    window_shape = [3]
     strides = [2]
     pads_begin = [0] * len(window_shape)
+    dilations = [1] * len(window_shape)
     pads_end = [0] * len(window_shape)
+    rounding_type = "floor"
+    auto_pad = "explicit"
+    idx_elem_type = "i32"
 
     model = ov.max_pool(
         A,
@@ -618,23 +560,22 @@ def test_max_pool():
         auto_pad,
         idx_elem_type,
     )
-    function = Model([model], parameter_list, "test")
 
-    size = 4
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(input_arr)[0]
+    assert model.get_type_name() == "MaxPool"
+    assert model.get_output_size() == 2
+    assert list(model.get_output_shape(0)) == [1, 1, 4]
+    assert list(model.get_output_shape(1)) == [1, 1, 4]
+    assert model.get_output_element_type(0) == element_type
+    assert model.get_output_element_type(1) == Type.i32
 
-    expected = ((np.arange(size) + 1) * 2).reshape(1, 1, size)
-    assert np.allclose(result, expected)
-
-    # test 2d
+def test_max_pool_2d():
     element_type = Type.f32
     shape = Shape([1, 1, 10, 10])
     A = Parameter(element_type, shape)
-    parameter_list = [A]
-
-    input_arr = np.arange(100, dtype=np.float32).reshape(1, 1, 10, 10)
     window_shape = [3, 3]
+    rounding_type = "floor"
+    auto_pad = "explicit"
+    idx_elem_type = "i32"
 
     strides = [1, 1]
     dilations = [1, 1]
@@ -652,19 +593,26 @@ def test_max_pool():
         auto_pad,
         idx_elem_type,
     )
-    function = Model([model], parameter_list, "test")
+    assert model.get_type_name() == "MaxPool"
+    assert model.get_output_size() == 2
+    assert list(model.get_output_shape(0)) == [1, 1, 8, 8]
+    assert list(model.get_output_shape(1)) == [1, 1, 8, 8]
+    assert model.get_output_element_type(0) == element_type
+    assert model.get_output_element_type(1) == Type.i32
 
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(input_arr)[0]
 
-    expected = ((np.arange(100).reshape(10, 10))[2:, 2:]).reshape(1, 1, 8, 8)
-    assert np.allclose(result, expected)
-
-    # test 2d with strides
+def test_max_pool_2d_with_strides():
+    element_type = Type.f32
+    shape = Shape([1, 1, 10, 10])
+    A = Parameter(element_type, shape)
     strides = [2, 2]
     dilations = [1, 1]
     pads_begin = [0, 0]
     pads_end = [0, 0]
+    window_shape = [3, 3]
+    rounding_type = "floor"
+    auto_pad = "explicit"
+    idx_elem_type = "i32"
 
     model = ov.max_pool(
         A,
@@ -677,13 +625,12 @@ def test_max_pool():
         auto_pad,
         idx_elem_type,
     )
-    function = Model([model], parameter_list, "test")
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(input_arr)[0]
-
-    size = 4
-    expected = ((np.arange(100).reshape(10, 10))[2::2, 2::2]).reshape(1, 1, size, size)
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "MaxPool"
+    assert model.get_output_size() == 2
+    assert list(model.get_output_shape(0)) == [1, 1, 4, 4]
+    assert list(model.get_output_shape(1)) == [1, 1, 4, 4]
+    assert model.get_output_element_type(0) == element_type
+    assert model.get_output_element_type(1) == Type.i32
 
 
 def convolution2d(
@@ -733,15 +680,11 @@ def convolution2d(
 
 
 def test_convolution_simple():
-
     element_type = Type.f32
     image_shape = Shape([1, 1, 16, 16])
     filter_shape = Shape([1, 1, 3, 3])
     data = Parameter(element_type, image_shape)
     filters = Parameter(element_type, filter_shape)
-    parameter_list = [data, filters]
-
-    image_arr = np.arange(-128, 128, 1, dtype=np.float32).reshape(1, 1, 16, 16)
     filter_arr = np.ones(9, dtype=np.float32).reshape(1, 1, 3, 3)
     filter_arr[0][0][0][0] = -1
     filter_arr[0][0][1][1] = -1
@@ -755,14 +698,11 @@ def test_convolution_simple():
     dilations = [1, 1]
 
     model = ov.convolution(data, filters, strides, pads_begin, pads_end, dilations)
-    function = Model([model], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(image_arr, filter_arr)[0]
-
-    expected = convolution2d(image_arr[0][0], filter_arr[0][0]).reshape(1, 1, 14, 14)
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Convolution"
+    assert model.get_output_size() == 1
+    assert list(model.get_output_shape(0)) == [1, 1, 14, 14]
+    assert model.get_output_element_type(0) == element_type
 
 
 def test_convolution_with_strides():
@@ -772,9 +712,6 @@ def test_convolution_with_strides():
     filter_shape = Shape([1, 1, 3, 3])
     data = Parameter(element_type, image_shape)
     filters = Parameter(element_type, filter_shape)
-    parameter_list = [data, filters]
-
-    image_arr = np.arange(100, dtype=np.float32).reshape([1, 1, 10, 10])
     filter_arr = np.zeros(9, dtype=np.float32).reshape([1, 1, 3, 3])
     filter_arr[0][0][1][1] = 1
     strides = [2, 2]
@@ -783,14 +720,11 @@ def test_convolution_with_strides():
     dilations = [1, 1]
 
     model = ov.convolution(data, filters, strides, pads_begin, pads_end, dilations)
-    function = Model([model], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(image_arr, filter_arr)[0]
-
-    expected = convolution2d(image_arr[0][0], filter_arr[0][0], strides).reshape(1, 1, 4, 4)
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Convolution"
+    assert model.get_output_size() == 1
+    assert list(model.get_output_shape(0)) == [1, 1, 4, 4]
+    assert model.get_output_element_type(0) == element_type
 
 
 def test_convolution_with_filter_dilation():
@@ -800,24 +734,17 @@ def test_convolution_with_filter_dilation():
     filter_shape = Shape([1, 1, 3, 3])
     data = Parameter(element_type, image_shape)
     filters = Parameter(element_type, filter_shape)
-    parameter_list = [data, filters]
-
-    image_arr = np.arange(100, dtype=np.float32).reshape([1, 1, 10, 10])
-    filter_arr = np.ones(9, dtype=np.float32).reshape([1, 1, 3, 3])
     strides = [1, 1]
     pads_begin = [0, 0]
     pads_end = [0, 0]
     dilations = [2, 2]
 
     model = ov.convolution(data, filters, strides, pads_begin, pads_end, dilations)
-    function = Model([model], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(image_arr, filter_arr)[0]
-
-    expected = convolution2d(image_arr[0][0], filter_arr[0][0], strides, dilations).reshape([1, 1, 6, 6])
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Convolution"
+    assert model.get_output_size() == 1
+    assert list(model.get_output_shape(0)) == [1, 1, 6, 6]
+    assert model.get_output_element_type(0) == element_type
 
 
 def test_convolution_with_padding():
@@ -827,9 +754,6 @@ def test_convolution_with_padding():
     filter_shape = Shape([1, 1, 3, 3])
     data = Parameter(element_type, image_shape)
     filters = Parameter(element_type, filter_shape)
-    parameter_list = [data, filters]
-
-    image_arr = np.arange(100, dtype=np.float32).reshape(1, 1, 10, 10)
     filter_arr = np.zeros(9, dtype=np.float32).reshape(1, 1, 3, 3)
     filter_arr[0][0][1][1] = 1
     strides = [1, 1]
@@ -838,16 +762,11 @@ def test_convolution_with_padding():
     pads_end = [0, 0]
 
     model = ov.convolution(data, filters, strides, pads_begin, pads_end, dilations)
-    function = Model([model], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(image_arr, filter_arr)[0]
-
-    expected = convolution2d(
-        image_arr[0][0], filter_arr[0][0], strides, dilations, pads_begin, pads_end
-    ).reshape([1, 1, 6, 6])
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Convolution"
+    assert model.get_output_size() == 1
+    assert list(model.get_output_shape(0)) == [1, 1, 6, 6]
+    assert model.get_output_element_type(0) == element_type
 
 
 def test_convolution_with_non_zero_padding():
@@ -856,9 +775,6 @@ def test_convolution_with_non_zero_padding():
     filter_shape = Shape([1, 1, 3, 3])
     data = Parameter(element_type, image_shape)
     filters = Parameter(element_type, filter_shape)
-    parameter_list = [data, filters]
-
-    image_arr = np.arange(100, dtype=np.float32).reshape(1, 1, 10, 10)
     filter_arr = (np.ones(9, dtype=np.float32).reshape(1, 1, 3, 3)) * -1
     filter_arr[0][0][1][1] = 1
     strides = [1, 1]
@@ -867,13 +783,8 @@ def test_convolution_with_non_zero_padding():
     pads_end = [1, 2]
 
     model = ov.convolution(data, filters, strides, pads_begin, pads_end, dilations)
-    function = Model([model], parameter_list, "test")
 
-    runtime = get_runtime()
-    computation = runtime.computation(function, *parameter_list)
-    result = computation(image_arr, filter_arr)[0]
-
-    expected = convolution2d(
-        image_arr[0][0], filter_arr[0][0], strides, dilations, pads_begin, pads_end
-    ).reshape([1, 1, 9, 9])
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Convolution"
+    assert model.get_output_size() == 1
+    assert list(model.get_output_shape(0)) == [1, 1, 9, 9]
+    assert model.get_output_element_type(0) == element_type
diff --git a/src/bindings/python/tests/test_graph/test_ops_binary.py b/src/bindings/python/tests/test_graph/test_ops_binary.py
index d27790641c3..d61fd62134b 100644
--- a/src/bindings/python/tests/test_graph/test_ops_binary.py
+++ b/src/bindings/python/tests/test_graph/test_ops_binary.py
@@ -7,28 +7,29 @@ import operator
 import numpy as np
 import pytest
 
+from openvino.runtime import Type
 import openvino.runtime.opset8 as ov
 
 
 @pytest.mark.parametrize(
-    ("graph_api_helper", "numpy_function"),
+    ("graph_api_helper", "numpy_function", "expected_type"),
     [
-        (ov.add, np.add),
-        (ov.divide, np.divide),
-        (ov.multiply, np.multiply),
-        (ov.subtract, np.subtract),
-        (ov.minimum, np.minimum),
-        (ov.maximum, np.maximum),
-        (ov.mod, np.mod),
-        (ov.equal, np.equal),
-        (ov.not_equal, np.not_equal),
-        (ov.greater, np.greater),
-        (ov.greater_equal, np.greater_equal),
-        (ov.less, np.less),
-        (ov.less_equal, np.less_equal),
+        (ov.add, np.add, Type.f32),
+        (ov.divide, np.divide, Type.f32),
+        (ov.multiply, np.multiply, Type.f32),
+        (ov.subtract, np.subtract, Type.f32),
+        (ov.minimum, np.minimum, Type.f32),
+        (ov.maximum, np.maximum, Type.f32),
+        (ov.mod, np.mod, Type.f32),
+        (ov.equal, np.equal, Type.boolean),
+        (ov.not_equal, np.not_equal, Type.boolean),
+        (ov.greater, np.greater, Type.boolean),
+        (ov.greater_equal, np.greater_equal, Type.boolean),
+        (ov.less, np.less, Type.boolean),
+        (ov.less_equal, np.less_equal, Type.boolean),
     ],
 )
-def test_binary_op(graph_api_helper, numpy_function):
+def test_binary_op(graph_api_helper, numpy_function, expected_type):
     shape = [2, 2]
     parameter_a = ov.parameter(shape, name="A", dtype=np.float32)
     parameter_b = ov.parameter(shape, name="B", dtype=np.float32)
@@ -41,27 +42,28 @@ def test_binary_op(graph_api_helper, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == expected_type
 
 
 @pytest.mark.parametrize(
-    ("graph_api_helper", "numpy_function"),
+    ("graph_api_helper", "numpy_function", "expected_type"),
     [
-        (ov.add, np.add),
-        (ov.divide, np.divide),
-        (ov.multiply, np.multiply),
-        (ov.subtract, np.subtract),
-        (ov.minimum, np.minimum),
-        (ov.maximum, np.maximum),
-        (ov.mod, np.mod),
-        (ov.equal, np.equal),
-        (ov.not_equal, np.not_equal),
-        (ov.greater, np.greater),
-        (ov.greater_equal, np.greater_equal),
-        (ov.less, np.less),
-        (ov.less_equal, np.less_equal),
+        (ov.add, np.add, Type.f32),
+        (ov.divide, np.divide, Type.f32),
+        (ov.multiply, np.multiply, Type.f32),
+        (ov.subtract, np.subtract, Type.f32),
+        (ov.minimum, np.minimum, Type.f32),
+        (ov.maximum, np.maximum, Type.f32),
+        (ov.mod, np.mod, Type.f32),
+        (ov.equal, np.equal, Type.boolean),
+        (ov.not_equal, np.not_equal, Type.boolean),
+        (ov.greater, np.greater, Type.boolean),
+        (ov.greater_equal, np.greater_equal, Type.boolean),
+        (ov.less, np.less, Type.boolean),
+        (ov.less_equal, np.less_equal, Type.boolean),
     ],
 )
-def test_binary_op_with_scalar(graph_api_helper, numpy_function):
+def test_binary_op_with_scalar(graph_api_helper, numpy_function, expected_type):
     value_a = np.array([[1, 2], [3, 4]], dtype=np.float32)
     value_b = np.array([[5, 6], [7, 8]], dtype=np.float32)
 
@@ -73,6 +75,7 @@ def test_binary_op_with_scalar(graph_api_helper, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == expected_type
 
 
 @pytest.mark.parametrize(
@@ -92,6 +95,7 @@ def test_binary_logical_op(graph_api_helper, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == Type.boolean
 
 
 @pytest.mark.parametrize(
@@ -110,24 +114,25 @@ def test_binary_logical_op_with_scalar(graph_api_helper, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == Type.boolean
 
 
 @pytest.mark.parametrize(
-    ("operator", "numpy_function"),
+    ("operator", "numpy_function", "expected_type"),
     [
-        (operator.add, np.add),
-        (operator.sub, np.subtract),
-        (operator.mul, np.multiply),
-        (operator.truediv, np.divide),
-        (operator.eq, np.equal),
-        (operator.ne, np.not_equal),
-        (operator.gt, np.greater),
-        (operator.ge, np.greater_equal),
-        (operator.lt, np.less),
-        (operator.le, np.less_equal),
+        (operator.add, np.add, Type.f32),
+        (operator.sub, np.subtract, Type.f32),
+        (operator.mul, np.multiply, Type.f32),
+        (operator.truediv, np.divide, Type.f32),
+        (operator.eq, np.equal, Type.boolean),
+        (operator.ne, np.not_equal, Type.boolean),
+        (operator.gt, np.greater, Type.boolean),
+        (operator.ge, np.greater_equal, Type.boolean),
+        (operator.lt, np.less, Type.boolean),
+        (operator.le, np.less_equal, Type.boolean),
     ],
 )
-def test_binary_operators(operator, numpy_function):
+def test_binary_operators(operator, numpy_function, expected_type):
     value_a = np.array([[1, 2], [3, 4]], dtype=np.float32)
     value_b = np.array([[4, 5], [1, 7]], dtype=np.float32)
 
@@ -139,24 +144,25 @@ def test_binary_operators(operator, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == expected_type
 
 
 @pytest.mark.parametrize(
-    ("operator", "numpy_function"),
+    ("operator", "numpy_function", "expected_type"),
     [
-        (operator.add, np.add),
-        (operator.sub, np.subtract),
-        (operator.mul, np.multiply),
-        (operator.truediv, np.divide),
-        (operator.eq, np.equal),
-        (operator.ne, np.not_equal),
-        (operator.gt, np.greater),
-        (operator.ge, np.greater_equal),
-        (operator.lt, np.less),
-        (operator.le, np.less_equal),
+        (operator.add, np.add, Type.f32),
+        (operator.sub, np.subtract, Type.f32),
+        (operator.mul, np.multiply, Type.f32),
+        (operator.truediv, np.divide, Type.f32),
+        (operator.eq, np.equal, Type.boolean),
+        (operator.ne, np.not_equal, Type.boolean),
+        (operator.gt, np.greater, Type.boolean),
+        (operator.ge, np.greater_equal, Type.boolean),
+        (operator.lt, np.less, Type.boolean),
+        (operator.le, np.less_equal, Type.boolean),
     ],
 )
-def test_binary_operators_with_scalar(operator, numpy_function):
+def test_binary_operators_with_scalar(operator, numpy_function, expected_type):
     value_a = np.array([[1, 2], [3, 4]], dtype=np.float32)
     value_b = np.array([[5, 6], [7, 8]], dtype=np.float32)
 
@@ -168,6 +174,7 @@ def test_binary_operators_with_scalar(operator, numpy_function):
     expected_shape = numpy_function(value_a, value_b).shape
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == list(expected_shape)
+    assert model.get_output_element_type(0) == expected_type
 
 
 def test_multiply():
@@ -180,6 +187,7 @@ def test_multiply():
     assert node.get_type_name() == "Multiply"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == list(expected_shape)
+    assert node.get_output_element_type(0) == Type.i32
 
 
 def test_power_v1():
@@ -192,3 +200,4 @@ def test_power_v1():
     assert node.get_type_name() == "Power"
     assert node.get_output_size() == 1
     assert list(node.get_output_shape(0)) == list(expected_shape)
+    assert node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_ops_fused.py b/src/bindings/python/tests/test_graph/test_ops_fused.py
index 19ce1dedfde..1e43d06df66 100644
--- a/src/bindings/python/tests/test_graph/test_ops_fused.py
+++ b/src/bindings/python/tests/test_graph/test_ops_fused.py
@@ -5,9 +5,8 @@
 import numpy as np
 import pytest
 
+import openvino.runtime as ov_runtime
 import openvino.runtime.opset8 as ov
-from tests.runtime import get_runtime
-from tests import xfail_issue_36486
 
 
 def test_elu_operator_with_scalar_and_array():
@@ -131,289 +130,126 @@ def test_squeeze_operator():
 
 
 def test_squared_difference_operator():
-    runtime = get_runtime()
-
     x1_shape = [1, 2, 3, 4]
     x2_shape = [2, 3, 4]
 
     parameter_x1 = ov.parameter(x1_shape, name="x1", dtype=np.float32)
     parameter_x2 = ov.parameter(x2_shape, name="x2", dtype=np.float32)
 
-    x1_value = np.arange(24.0, dtype=np.float32).reshape(x1_shape)
-    x2_value = np.arange(start=4.0, stop=28.0, step=1.0, dtype=np.float32).reshape(x2_shape)
-
     model = ov.squared_difference(parameter_x1, parameter_x2)
-    computation = runtime.computation(model, parameter_x1, parameter_x2)
-
-    result = computation(x1_value, x2_value)
-    expected = np.square(np.subtract(x1_value, x2_value))
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "SquaredDifference"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 2, 3, 4]
 
 
 def test_shuffle_channels_operator():
-    runtime = get_runtime()
-
     data_shape = [1, 15, 2, 2]
     axis = 1
     groups = 5
-
     parameter = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
-    data_value = np.arange(60.0, dtype=np.float32).reshape(data_shape)
-
     model = ov.shuffle_channels(parameter, axis, groups)
-    computation = runtime.computation(model, parameter)
-
-    result = computation(data_value)
-    expected = np.array(
-        [
-            [
-                [[0.0, 1.0], [2.0, 3.0]],
-                [[12.0, 13.0], [14.0, 15.0]],
-                [[24.0, 25.0], [26.0, 27.0]],
-                [[36.0, 37.0], [38.0, 39.0]],
-                [[48.0, 49.0], [50.0, 51.0]],
-                [[4.0, 5.0], [6.0, 7.0]],
-                [[16.0, 17.0], [18.0, 19.0]],
-                [[28.0, 29.0], [30.0, 31.0]],
-                [[40.0, 41.0], [42.0, 43.0]],
-                [[52.0, 53.0], [54.0, 55.0]],
-                [[8.0, 9.0], [10.0, 11.0]],
-                [[20.0, 21.0], [22.0, 23.0]],
-                [[32.0, 33.0], [34.0, 35.0]],
-                [[44.0, 45.0], [46.0, 47.0]],
-                [[56.0, 57.0], [58.0, 59.0]],
-            ],
-        ],
-        dtype=np.float32,
-    )
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "ShuffleChannels"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 15, 2, 2]
 
 
 def test_unsqueeze():
-    runtime = get_runtime()
-
     data_shape = [3, 4, 5]
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
-    data_value = np.arange(60.0, dtype=np.float32).reshape(3, 4, 5)
     axes = [0, 4]
-    model = ov.unsqueeze(parameter_data, axes)
-    computation = runtime.computation(model, parameter_data)
 
-    result = computation(data_value)
-    expected = np.arange(60.0, dtype=np.float32).reshape([1, 3, 4, 5, 1])
-    assert np.allclose(result, expected)
+    model = ov.unsqueeze(parameter_data, axes)
+    assert model.get_type_name() == "Unsqueeze"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 3, 4, 5, 1]
 
 
 def test_grn_operator():
-    runtime = get_runtime()
-
-    data_value = np.arange(start=1.0, stop=25.0, dtype=np.float32).reshape([1, 2, 3, 4])
     bias = np.float32(1e-6)
-
     data_shape = [1, 2, 3, 4]
-
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
     model = ov.grn(parameter_data, bias)
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data_value)
-    expected = np.array(
-        [
-            [
-                [
-                    [0.0766965, 0.14142136, 0.19611613, 0.24253564],
-                    [0.28216633, 0.31622776, 0.34570536, 0.37139067],
-                    [0.39391932, 0.41380295, 0.4314555, 0.4472136],
-                ],
-                [
-                    [0.9970545, 0.98994946, 0.9805807, 0.97014254],
-                    [0.9593655, 0.9486833, 0.9383431, 0.9284767],
-                    [0.91914505, 0.9103665, 0.9021342, 0.8944272],
-                ],
-            ],
-        ],
-        dtype=np.float32,
-    )
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "GRN"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 2, 3, 4]
 
 
 def test_prelu_operator():
-    runtime = get_runtime()
-
     data_shape = [1, 2, 3, 4]
     slope_shape = [2, 3, 1]
-
     data_value = np.arange(start=1.0, stop=25.0, dtype=np.float32).reshape(data_shape)
     slope_value = np.arange(start=-10.0, stop=-4.0, dtype=np.float32).reshape(slope_shape)
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
     parameter_slope = ov.parameter(slope_shape, name="Slope", dtype=np.float32)
 
     model = ov.prelu(parameter_data, parameter_slope)
-    computation = runtime.computation(model, parameter_data, parameter_slope)
-
-    result = computation(data_value, slope_value)
     expected = np.clip(data_value, 0, np.inf) + np.clip(data_value, -np.inf, 0) * slope_value
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "PRelu"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == list(expected.shape)
 
 
 def test_selu_operator():
-    runtime = get_runtime()
-
     data_shape = [4, 2, 3, 1]
-
-    data = np.arange(start=-1.0, stop=23.0, dtype=np.float32).reshape(data_shape)
     alpha = np.array(1.6733, dtype=np.float32)
     lambda_value = np.array(1.0507, dtype=np.float32)
 
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
     model = ov.selu(parameter_data, alpha, lambda_value)
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data)
-    mask = (data > 0) * data + (data <= 0) * (alpha * np.exp(data) - alpha)
-    expected = mask * lambda_value
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "Selu"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [4, 2, 3, 1]
 
 
-@xfail_issue_36486
 def test_hard_sigmoid_operator():
-    runtime = get_runtime()
-
     data_shape = [3]
-    alpha_value = np.float32(0.5)
-    beta_value = np.float32(0.6)
-
-    data_value = np.array([-1, 0, 1], dtype=np.float32)
-
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
     parameter_alpha = ov.parameter([], name="Alpha", dtype=np.float32)
     parameter_beta = ov.parameter([], name="Beta", dtype=np.float32)
 
     model = ov.hard_sigmoid(parameter_data, parameter_alpha, parameter_beta)
-    computation = runtime.computation(model, parameter_data, parameter_alpha, parameter_beta)
-
-    result = computation(data_value, alpha_value, beta_value)
-    expected = [0.1, 0.6, 1.0]
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "HardSigmoid"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [3]
 
 
 def test_mvn_operator():
-    runtime = get_runtime()
-
     data_shape = [3, 3, 3, 1]
     axes = [0, 2, 3]
     normalize_variance = True
     eps = np.float32(1e-9)
     eps_mode = "outside_sqrt"
-
-    data_value = np.array(
-        [
-            [
-                [[0.8439683], [0.5665144], [0.05836735]],
-                [[0.02916367], [0.12964272], [0.5060197]],
-                [[0.79538304], [0.9411346], [0.9546573]],
-            ],
-            [
-                [[0.17730942], [0.46192095], [0.26480448]],
-                [[0.6746842], [0.01665257], [0.62473077]],
-                [[0.9240844], [0.9722341], [0.11965699]],
-            ],
-            [
-                [[0.41356155], [0.9129373], [0.59330076]],
-                [[0.81929934], [0.7862604], [0.11799799]],
-                [[0.69248444], [0.54119414], [0.07513223]],
-            ],
-        ],
-        dtype=np.float32,
-    )
-
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
     model = ov.mvn(parameter_data, axes, normalize_variance, eps, eps_mode)
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data_value)
-
-    expected = np.array(
-        [
-            [
-                [[1.3546423], [0.33053496], [-1.5450814]],
-                [[-1.2106764], [-0.8925952], [0.29888135]],
-                [[0.38083088], [0.81808794], [0.85865635]],
-            ],
-            [
-                [[-1.1060555], [-0.05552877], [-0.78310335]],
-                [[0.83281356], [-1.250282], [0.67467856]],
-                [[0.7669372], [0.9113869], [-1.6463585]],
-            ],
-            [
-                [[-0.23402764], [1.6092131], [0.42940593]],
-                [[1.2906139], [1.1860244], [-0.92945826]],
-                [[0.0721334], [-0.38174], [-1.7799333]],
-            ],
-        ],
-        dtype=np.float32,
-    )
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "MVN"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [3, 3, 3, 1]
 
 
 @pytest.mark.skip(reason="Sporadically failed. Need further investigation. Ticket - 95970")
 def test_space_to_depth_operator():
-    runtime = get_runtime()
-
     data_shape = [1, 2, 4, 4]
-    data_value = np.arange(start=0, stop=32, step=1.0, dtype=np.float32).reshape(data_shape)
     mode = "blocks_first"
     block_size = 2
 
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
     model = ov.space_to_depth(parameter_data, mode, block_size)
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data_value)
-    expected = np.array(
-        [
-            0,
-            2,
-            8,
-            10,
-            16,
-            18,
-            24,
-            26,
-            1,
-            3,
-            9,
-            11,
-            17,
-            19,
-            25,
-            27,
-            4,
-            6,
-            12,
-            14,
-            20,
-            22,
-            28,
-            30,
-            5,
-            7,
-            13,
-            15,
-            21,
-            23,
-            29,
-            31,
-        ],
-        dtype=np.float32,
-    ).reshape(1, 8, 2, 2)
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "SpaceToDepth"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 8, 2, 2]
 
     batch_size = 2
     input_size = 3
@@ -431,41 +267,6 @@ def test_space_to_depth_operator():
     parameter_r = ov.parameter(r_shape, name="R", dtype=np.float32)
     parameter_b = ov.parameter(b_shape, name="B", dtype=np.float32)
 
-    x_value = np.array(
-        [0.3432185, 0.612268, 0.20272376, 0.9513413, 0.30585995, 0.7265472], dtype=np.float32,
-    ).reshape(x_shape)
-    h_t_value = np.array(
-        [0.12444675, 0.52055854, 0.46489045, 0.4983964, 0.7730452, 0.28439692], dtype=np.float32,
-    ).reshape(h_t_shape)
-    w_value = np.array(
-        [
-            0.41930267,
-            0.7872176,
-            0.89940447,
-            0.23659843,
-            0.24676207,
-            0.17101714,
-            0.3147149,
-            0.6555601,
-            0.4559603,
-        ],
-        dtype=np.float32,
-    ).reshape(w_shape)
-    r_value = np.array(
-        [
-            0.8374871,
-            0.86660194,
-            0.82114047,
-            0.71549815,
-            0.18775631,
-            0.3182116,
-            0.25392973,
-            0.38301638,
-            0.85531586,
-        ],
-        dtype=np.float32,
-    ).reshape(r_shape)
-    b_value = np.array([1.0289404, 1.6362579, 0.4370661], dtype=np.float32).reshape(b_shape)
     activations = ["sigmoid"]
     activation_alpha = []
     activation_beta = []
@@ -483,47 +284,33 @@ def test_space_to_depth_operator():
         activation_beta,
         clip,
     )
-    computation = runtime.computation(
-        model, parameter_x, parameter_h_t, parameter_w, parameter_r, parameter_b,
-    )
-
-    result = computation(x_value, h_t_value, w_value, r_value, b_value)
-    expected = np.array(
-        [0.94126844, 0.9036043, 0.841243, 0.9468489, 0.934215, 0.873708], dtype=np.float32,
-    ).reshape(batch_size, hidden_size)
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "SpaceToDepth"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [batch_size, hidden_size]
 
 
 def test_group_convolution_operator():
-    runtime = get_runtime()
-
     data_shape = [1, 4, 2, 2]
     filters_shape = [2, 1, 2, 1, 1]
 
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
     parameter_filters = ov.parameter(filters_shape, name="Filters", dtype=np.float32)
 
-    data_value = np.arange(start=1.0, stop=17.0, dtype=np.float32).reshape(data_shape)
-    filters_value = np.arange(start=1.0, stop=5.0, dtype=np.float32).reshape(filters_shape)
     strides = [1, 1]
     dilations = [1, 1]
     pads_begin = [0, 0]
     pads_end = [0, 0]
 
     model = ov.group_convolution(parameter_data, parameter_filters, strides, pads_begin, pads_end, dilations)
-    computation = runtime.computation(model, parameter_data, parameter_filters)
-    result = computation(data_value, filters_value)
-
-    expected = np.array([11, 14, 17, 20, 79, 86, 93, 100], dtype=np.float32).reshape(1, 2, 2, 2)
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "GroupConvolution"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 2, 2, 2]
 
 
 @pytest.mark.xfail(reason="Computation mismatch")
 def test_group_convolution_backprop_data():
-    runtime = get_runtime()
-
     data_shape = [1, 1, 3, 3]
     filters_shape = [1, 1, 1, 3, 3]
     strides = [2, 2]
@@ -537,87 +324,13 @@ def test_group_convolution_backprop_data():
         data_node, filters_node, strides, None, pads_begin, pads_end, output_padding=output_padding,
     )
 
-    data_value = np.array(
-        [
-            0.16857791,
-            -0.15161794,
-            0.08540368,
-            0.1820628,
-            -0.21746576,
-            0.08245695,
-            0.1431433,
-            -0.43156421,
-            0.30591947,
-        ],
-        dtype=np.float32,
-    ).reshape(data_shape)
-
-    filters_value = np.array(
-        [
-            -0.06230065,
-            0.37932432,
-            -0.25388849,
-            0.33878803,
-            0.43709868,
-            -0.22477469,
-            0.04118127,
-            -0.44696793,
-            0.06373066,
-        ],
-        dtype=np.float32,
-    ).reshape(filters_shape)
-
-    computation = runtime.computation(model, data_node, filters_node)
-    result = computation(data_value, filters_value)
-
-    expected = np.array(
-        [
-            0.07368518,
-            -0.08925839,
-            -0.06627201,
-            0.06301362,
-            0.03732984,
-            -0.01919658,
-            -0.00628807,
-            -0.02817563,
-            -0.01472169,
-            0.04392925,
-            -0.00689478,
-            -0.01549204,
-            0.07957941,
-            -0.11459791,
-            -0.09505399,
-            0.07681622,
-            0.03604182,
-            -0.01853423,
-            -0.0270785,
-            -0.00680824,
-            -0.06650258,
-            0.08004665,
-            0.07918708,
-            0.0724144,
-            0.06256775,
-            -0.17838378,
-            -0.18863615,
-            0.20064656,
-            0.133717,
-            -0.06876295,
-            -0.06398046,
-            -0.00864975,
-            0.19289537,
-            -0.01490572,
-            -0.13673618,
-            0.01949645,
-        ],
-        dtype=np.float32,
-    ).reshape(1, 1, 6, 6)
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "GroupConvolutionBackpropData"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 1, 6, 6]
 
 
 def test_group_convolution_backprop_data_output_shape():
-    runtime = get_runtime()
-
     data_shape = [1, 1, 1, 10]
     filters_shape = [1, 1, 1, 1, 5]
     strides = [1, 1]
@@ -630,17 +343,7 @@ def test_group_convolution_backprop_data_output_shape():
         data_node, filters_node, strides, output_shape_node, auto_pad="same_upper",
     )
 
-    data_value = np.array([0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0], dtype=np.float32).reshape(
-        data_shape,
-    )
-
-    filters_value = np.array([1.0, 2.0, 3.0, 2.0, 1.0], dtype=np.float32).reshape(filters_shape)
-
-    computation = runtime.computation(model, data_node, filters_node)
-    result = computation(data_value, filters_value)
-
-    expected = np.array(
-        [0.0, 1.0, 4.0, 10.0, 18.0, 27.0, 36.0, 45.0, 54.0, 63.0, 62.0, 50.0, 26.0, 9.0], dtype=np.float32,
-    ).reshape(1, 1, 1, 14)
-
-    assert np.allclose(result, expected)
+    assert model.get_type_name() == "GroupConvolutionBackpropData"
+    assert model.get_output_size() == 1
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [1, 1, 1, 14]
diff --git a/src/bindings/python/tests/test_graph/test_ops_matmul.py b/src/bindings/python/tests/test_graph/test_ops_matmul.py
index 3d899831331..cb6abbdd773 100644
--- a/src/bindings/python/tests/test_graph/test_ops_matmul.py
+++ b/src/bindings/python/tests/test_graph/test_ops_matmul.py
@@ -6,37 +6,31 @@ import numpy as np
 import pytest
 
 import openvino.runtime.opset8 as ov
-from tests.test_graph.util import run_op_node
 
 
 @pytest.mark.parametrize(
-    ("shape_a", "shape_b", "transpose_a", "transpose_b"),
+    ("shape_a", "shape_b", "transpose_a", "transpose_b", "output_shape"),
     [
         # matrix, vector
-        ([2, 4], [4], False, False),
-        ([4], [4, 2], False, False),
+        ([2, 4], [4], False, False, [2]),
+        ([4], [4, 2], False, False, [2]),
         # matrix, matrix
-        ([2, 4], [4, 2], False, False),
+        ([2, 4], [4, 2], False, False, [2, 2]),
         # tensor, vector
-        ([2, 4, 5], [5], False, False),
+        ([2, 4, 5], [5], False, False, [2, 4]),
         # # tensor, matrix
-        ([2, 4, 5], [5, 4], False, False),
+        ([2, 4, 5], [5, 4], False, False, [2, 4, 4]),
         # # tensor, tensor
-        ([2, 2, 4], [2, 4, 2], False, False),
+        ([2, 2, 4], [2, 4, 2], False, False, [2, 2, 2]),
     ],
 )
 @pytest.mark.skip(reason="Sporadically failed. Need further investigation. Ticket - 95970")
-def test_matmul(shape_a, shape_b, transpose_a, transpose_b):
+def test_matmul(shape_a, shape_b, transpose_a, transpose_b, output_shape):
     np.random.seed(133391)
     left_input = -100.0 + np.random.rand(*shape_a).astype(np.float32) * 200.0
     right_input = -100.0 + np.random.rand(*shape_b).astype(np.float32) * 200.0
 
-    result = run_op_node([left_input, right_input], ov.matmul, transpose_a, transpose_b)
-
-    if transpose_a:
-        left_input = np.transpose(left_input)
-    if transpose_b:
-        right_input = np.transpose(right_input)
-
-    expected = np.matmul(left_input, right_input)
-    assert np.allclose(result, expected)
+    node = ov.matmul(left_input, right_input, transpose_a, transpose_b)
+    assert node.get_type_name() == "MatMul"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == output_shape
diff --git a/src/bindings/python/tests/test_graph/test_ops_multioutput.py b/src/bindings/python/tests/test_graph/test_ops_multioutput.py
index 1bf4c9ac6be..2650a7b6fa6 100644
--- a/src/bindings/python/tests/test_graph/test_ops_multioutput.py
+++ b/src/bindings/python/tests/test_graph/test_ops_multioutput.py
@@ -5,33 +5,28 @@
 import numpy as np
 
 import openvino.runtime.opset8 as ov
-from tests.runtime import get_runtime
 
 
 def test_split():
-    runtime = get_runtime()
     input_tensor = ov.constant(np.array([0, 1, 2, 3, 4, 5], dtype=np.int32))
     axis = ov.constant(0, dtype=np.int64)
     splits = 3
 
     split_node = ov.split(input_tensor, axis, splits)
-    computation = runtime.computation(split_node)
-    split_results = computation()
-    expected_results = np.array([[0, 1], [2, 3], [4, 5]], dtype=np.int32)
-    assert np.allclose(split_results, expected_results)
+    assert split_node.get_type_name() == "Split"
+    assert split_node.get_output_size() == 3
+    assert list(split_node.get_output_shape(0)) == [2]
+    assert list(split_node.get_output_shape(1)) == [2]
+    assert list(split_node.get_output_shape(2)) == [2]
 
 
 def test_variadic_split():
-    runtime = get_runtime()
     input_tensor = ov.constant(np.array([[0, 1, 2, 3, 4, 5], [6, 7, 8, 9, 10, 11]], dtype=np.int32))
     axis = ov.constant(1, dtype=np.int64)
     splits = ov.constant(np.array([2, 4], dtype=np.int64))
 
     v_split_node = ov.variadic_split(input_tensor, axis, splits)
-    computation = runtime.computation(v_split_node)
-    results = computation()
-    split0 = np.array([[0, 1], [6, 7]], dtype=np.int32)
-    split1 = np.array([[2, 3, 4, 5], [8, 9, 10, 11]], dtype=np.int32)
-
-    assert np.allclose(results[0], split0)
-    assert np.allclose(results[1], split1)
+    assert v_split_node.get_type_name() == "VariadicSplit"
+    assert v_split_node.get_output_size() == 2
+    assert list(v_split_node.get_output_shape(0)) == [2, 2]
+    assert list(v_split_node.get_output_shape(1)) == [2, 4]
diff --git a/src/bindings/python/tests/test_graph/test_ops_reshape.py b/src/bindings/python/tests/test_graph/test_ops_reshape.py
index f09c6d6ffc4..fe5520c0800 100644
--- a/src/bindings/python/tests/test_graph/test_ops_reshape.py
+++ b/src/bindings/python/tests/test_graph/test_ops_reshape.py
@@ -2,36 +2,37 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
+import openvino.runtime as ov_runtime
 import openvino.runtime.opset8 as ov
 import numpy as np
 import pytest
 
-from tests.runtime import get_runtime
-from tests.test_graph.util import run_op_node, run_op_numeric_data
+from openvino.runtime.utils.types import get_element_type
 
 
 def test_concat():
     input_a = np.array([[1, 2], [3, 4]]).astype(np.float32)
     input_b = np.array([[5, 6]]).astype(np.float32)
     axis = 0
-    expected = np.concatenate((input_a, input_b), axis=0)
+    expected_shape = np.concatenate((input_a, input_b), axis=0).shape
 
-    runtime = get_runtime()
     parameter_a = ov.parameter(list(input_a.shape), name="A", dtype=np.float32)
     parameter_b = ov.parameter(list(input_b.shape), name="B", dtype=np.float32)
     node = ov.concat([parameter_a, parameter_b], axis)
-    computation = runtime.computation(node, parameter_a, parameter_b)
-    result = computation(input_a, input_b)
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == "Concat"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == list(expected_shape)
 
 
 @pytest.mark.parametrize(
-    ("val_type", "value"), [(bool, False), (bool, np.empty((2, 2), dtype=bool))],
+    ("val_type", "value", "output_shape"), [(bool, False, []), (bool, np.empty((2, 2), dtype=bool), [2, 2])],
 )
-def test_constant_from_bool(val_type, value):
-    expected = np.array(value, dtype=val_type)
-    result = run_op_numeric_data(value, ov.constant, val_type)
-    assert np.allclose(result, expected)
+def test_constant_from_bool(val_type, value, output_shape):
+    node = ov.constant(value, val_type)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.boolean
+    assert list(node.get_output_shape(0)) == output_shape
 
 
 @pytest.mark.parametrize(
@@ -50,9 +51,11 @@ def test_constant_from_bool(val_type, value):
     ],
 )
 def test_constant_from_scalar(val_type, value):
-    expected = np.array(value, dtype=val_type)
-    result = run_op_numeric_data(value, ov.constant, val_type)
-    assert np.allclose(result, expected)
+    node = ov.constant(value, val_type)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(val_type)
+    assert list(node.get_output_shape(0)) == []
 
 
 @pytest.mark.parametrize(
@@ -65,8 +68,11 @@ def test_constant_from_scalar(val_type, value):
 def test_constant_from_float_array(val_type):
     np.random.seed(133391)
     input_data = np.array(-1 + np.random.rand(2, 3, 4) * 2, dtype=val_type)
-    result = run_op_numeric_data(input_data, ov.constant, val_type)
-    assert np.allclose(result, input_data)
+    node = ov.constant(input_data, val_type)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(val_type)
+    assert list(node.get_output_shape(0)) == [2, 3, 4]
 
 
 @pytest.mark.parametrize(
@@ -87,8 +93,11 @@ def test_constant_from_integer_array(val_type, range_start, range_end):
     input_data = np.array(
         np.random.randint(range_start, range_end, size=(2, 2)), dtype=val_type,
     )
-    result = run_op_numeric_data(input_data, ov.constant, val_type)
-    assert np.allclose(result, input_data)
+    node = ov.constant(input_data, val_type)
+    assert node.get_type_name() == "Constant"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == get_element_type(val_type)
+    assert list(node.get_output_shape(0)) == [2, 2]
 
 
 def test_broadcast_numpy():
@@ -127,27 +136,24 @@ def test_transpose():
     )
     input_order = np.array([0, 2, 3, 1], dtype=np.int32)
 
-    result = run_op_node([input_tensor], ov.transpose, input_order)
-
-    expected = np.transpose(input_tensor, input_order)
-
-    assert np.allclose(result, expected)
+    node = ov.transpose(input_tensor, input_order)
+    assert node.get_type_name() == "Transpose"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.i32
+    assert list(node.get_output_shape(0)) == [3, 224, 224, 3]
 
 
 def test_tile():
     input_tensor = np.arange(6, dtype=np.int32).reshape((2, 1, 3))
     repeats = np.array([2, 1], dtype=np.int32)
+    node = ov.tile(input_tensor, repeats)
 
-    result = run_op_node([input_tensor], ov.tile, repeats)
-
-    expected = np.array([0, 1, 2, 0, 1, 2, 3, 4, 5, 3, 4, 5]).reshape((2, 2, 3))
-
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == "Tile"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.i32
+    assert list(node.get_output_shape(0)) == [2, 2, 3]
 
 
-@pytest.mark.xfail(
-    reason="RuntimeError: Check 'shape_size(get_input_shape(0)) == shape_size(output_shape)'",
-)
 def test_strided_slice():
     input_tensor = np.arange(2 * 3 * 4, dtype=np.float32).reshape((2, 3, 4))
     begin = np.array([1, 0], dtype=np.int32)
@@ -159,9 +165,8 @@ def test_strided_slice():
     shrink_axis_mask = np.array([1, 0, 0], dtype=np.int32)
     ellipsis_mask = np.array([0, 0, 0], dtype=np.int32)
 
-    result = run_op_node(
-        [input_tensor],
-        ov.strided_slice,
+    node = ov.strided_slice(
+        input_tensor,
         begin,
         end,
         strides,
@@ -171,12 +176,10 @@ def test_strided_slice():
         shrink_axis_mask,
         ellipsis_mask,
     )
-
-    expected = np.array(
-        [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23], dtype=np.float32,
-    ).reshape((1, 3, 4))
-
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == "StridedSlice"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [1, 3, 4]
 
 
 def test_reshape_v1():
@@ -184,16 +187,18 @@ def test_reshape_v1():
     shape = np.array([0, -1, 4], dtype=np.int32)
     special_zero = True
 
-    expected_shape = np.array([2, 150, 4])
-    expected = np.reshape(param_a, expected_shape)
-    result = run_op_node([param_a], ov.reshape, shape, special_zero)
-
-    assert np.allclose(result, expected)
+    node = ov.reshape(param_a, shape, special_zero)
+    assert node.get_type_name() == "Reshape"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [2, 150, 4]
 
 
 def test_shape_of():
     input_tensor = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]], dtype=np.float32)
 
-    result = run_op_node([input_tensor], ov.shape_of)
-
-    assert np.allclose(result, [3, 3])
+    node = ov.shape_of(input_tensor)
+    assert node.get_type_name() == "ShapeOf"
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.i64
+    assert list(node.get_output_shape(0)) == [2]
diff --git a/src/bindings/python/tests/test_graph/test_ops_unary.py b/src/bindings/python/tests/test_graph/test_ops_unary.py
index 8a006d4e83d..a63bd843251 100644
--- a/src/bindings/python/tests/test_graph/test_ops_unary.py
+++ b/src/bindings/python/tests/test_graph/test_ops_unary.py
@@ -5,48 +5,49 @@
 import numpy as np
 import pytest
 
+import openvino.runtime as ov_runtime
 import openvino.runtime.opset9 as ov
 from openvino.runtime import Shape, Type
-from tests.runtime import get_runtime
-from tests.test_graph.util import run_op_node
 
 R_TOLERANCE = 1e-6  # global relative tolerance
 
 
 @pytest.mark.parametrize(
-    ("graph_api_fn", "numpy_fn", "range_start", "range_end"),
+    ("graph_api_fn", "type_name"),
     [
-        (ov.absolute, np.abs, -1, 1),
-        (ov.abs, np.abs, -1, 1),
-        (ov.acos, np.arccos, -1, 1),
-        (ov.acosh, np.arccosh, 1, 2),
-        (ov.asin, np.arcsin, -1, 1),
-        (ov.asinh, np.arcsinh, -1, 1),
-        (ov.atan, np.arctan, -100.0, 100.0),
-        (ov.atanh, np.arctanh, 0.0, 1.0),
-        (ov.ceiling, np.ceil, -100.0, 100.0),
-        (ov.ceil, np.ceil, -100.0, 100.0),
-        (ov.cos, np.cos, -100.0, 100.0),
-        (ov.cosh, np.cosh, -100.0, 100.0),
-        (ov.exp, np.exp, -100.0, 100.0),
-        (ov.floor, np.floor, -100.0, 100.0),
-        (ov.log, np.log, 0, 100.0),
-        (ov.relu, lambda x: np.maximum(0, x), -100.0, 100.0),
-        (ov.sign, np.sign, -100.0, 100.0),
-        (ov.sin, np.sin, -100.0, 100.0),
-        (ov.sinh, np.sinh, -100.0, 100.0),
-        (ov.sqrt, np.sqrt, 0.0, 100.0),
-        (ov.tan, np.tan, -1.0, 1.0),
-        (ov.tanh, np.tanh, -100.0, 100.0),
+        (ov.absolute, "Abs"),
+        (ov.abs, "Abs"),
+        (ov.acos, "Acos"),
+        (ov.acosh, "Acosh"),
+        (ov.asin, "Asin"),
+        (ov.asinh, "Asinh"),
+        (ov.atan, "Atan"),
+        (ov.atanh, "Atanh"),
+        (ov.ceiling, "Ceiling"),
+        (ov.ceil, "Ceiling"),
+        (ov.cos, "Cos"),
+        (ov.cosh, "Cosh"),
+        (ov.exp, "Exp"),
+        (ov.floor, "Floor"),
+        (ov.log, "Log"),
+        (ov.relu, "Relu"),
+        (ov.sign, "Sign"),
+        (ov.sin, "Sin"),
+        (ov.sinh, "Sinh"),
+        (ov.sqrt, "Sqrt"),
+        (ov.tan, "Tan"),
+        (ov.tanh, "Tanh"),
     ],
 )
-def test_unary_op_array(graph_api_fn, numpy_fn, range_start, range_end):
+def test_unary_op_array(graph_api_fn, type_name):
     np.random.seed(133391)
-    input_data = (range_start + np.random.rand(2, 3, 4) * (range_end - range_start)).astype(np.float32)
-    expected = numpy_fn(input_data)
+    input_data = np.random.rand(2, 3, 4).astype(np.float32)
 
-    result = run_op_node([input_data], graph_api_fn)
-    assert np.allclose(result, expected, rtol=0.001)
+    node = graph_api_fn(input_data)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == type_name
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [2, 3, 4]
 
 
 @pytest.mark.parametrize(
@@ -74,10 +75,12 @@ def test_unary_op_array(graph_api_fn, numpy_fn, range_start, range_end):
     ],
 )
 def test_unary_op_scalar(graph_api_fn, numpy_fn, input_data):
-    expected = numpy_fn(input_data)
+    expected_shape = numpy_fn(input_data).shape
+    node = graph_api_fn(input_data)
 
-    result = run_op_node([input_data], graph_api_fn)
-    assert np.allclose(result, expected)
+    assert node.get_output_size() == 1
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == list(expected_shape)
 
 
 @pytest.mark.parametrize(
@@ -85,52 +88,42 @@ def test_unary_op_scalar(graph_api_fn, numpy_fn, input_data):
     [(np.array([True, False, True, False])), (np.array([True])), (np.array([False]))],
 )
 def test_logical_not(input_data):
-    expected = np.logical_not(input_data)
-
-    result = run_op_node([input_data], ov.logical_not)
-    assert np.allclose(result, expected)
+    node = ov.logical_not(input_data)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "LogicalNot"
+    assert node.get_output_element_type(0) == ov_runtime.Type.boolean
+    assert list(node.get_output_shape(0)) == list(input_data.shape)
 
 
 def test_sigmoid():
     input_data = np.array([-3.14, -1.0, 0.0, 2.71001, 1000.0], dtype=np.float32)
-    result = run_op_node([input_data], ov.sigmoid)
+    node = ov.sigmoid(input_data)
 
-    def sigmoid(value):
-        return 1.0 / (1.0 + np.exp(-value))
-
-    expected = np.array(list(map(sigmoid, input_data)))
-
-    assert np.allclose(result, expected)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "Sigmoid"
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [5]
 
 
-def test_softmax_positive_axis():
+def test_softmax():
     axis = 1
     input_tensor = np.array([[1, 2, 3], [4, 5, 6]], dtype=np.float32)
 
-    result = run_op_node([input_tensor], ov.softmax, axis)
-
-    expected = [[0.09003056, 0.24472842, 0.6652409], [0.09003056, 0.24472842, 0.6652409]]
-
-    assert np.allclose(result, expected)
-
-
-def test_softmax_negative_axis():
-    axis = -1
-    input_tensor = np.array([[1, 2, 3], [4, 5, 6]], dtype=np.float32)
-
-    result = run_op_node([input_tensor], ov.softmax, axis)
-
-    expected = [[0.09003056, 0.24472842, 0.6652409], [0.09003056, 0.24472842, 0.6652409]]
-
-    assert np.allclose(result, expected)
+    node = ov.softmax(input_tensor, axis)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "Softmax"
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [2, 3]
 
 
 def test_erf():
     input_tensor = np.array([-1.0, 0.0, 1.0, 2.5, 3.14, 4.0], dtype=np.float32)
-    expected = [-0.842701, 0.0, 0.842701, 0.999593, 0.999991, 1.0]
 
-    result = run_op_node([input_tensor], ov.erf)
-    assert np.allclose(result, expected)
+    node = ov.erf(input_tensor)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "Erf"
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [6]
 
 
 def test_hswish():
@@ -144,7 +137,7 @@ def test_hswish():
     assert node.get_output_element_type(0) == Type.f32
 
 
-def test_round_even():
+def test_round():
     float_dtype = np.float32
     data = ov.parameter(Shape([3, 10]), dtype=float_dtype, name="data")
 
@@ -155,27 +148,12 @@ def test_round_even():
     assert node.get_output_element_type(0) == Type.f32
 
     input_tensor = np.array([-2.5, -1.5, -0.5, 0.5, 0.9, 1.5, 2.3, 2.5, 3.5], dtype=np.float32)
-    expected = [-2.0, -2.0, 0.0, 0.0, 1.0, 2.0, 2.0, 2.0, 4.0]
 
-    result = run_op_node([input_tensor], ov.round, "HALF_TO_EVEN")
-    assert np.allclose(result, expected)
-
-
-def test_round_away():
-    float_dtype = np.float32
-    data = ov.parameter(Shape([3, 10]), dtype=float_dtype, name="data")
-
-    node = ov.round(data, "HALF_AWAY_FROM_ZERO")
-    assert node.get_type_name() == "Round"
+    node = ov.round(input_tensor, "HALF_TO_EVEN")
     assert node.get_output_size() == 1
-    assert list(node.get_output_shape(0)) == [3, 10]
-    assert node.get_output_element_type(0) == Type.f32
-
-    input_tensor = np.array([-2.5, -1.5, -0.5, 0.5, 0.9, 1.5, 2.3, 2.5, 3.5], dtype=np.float32)
-    expected = [-3.0, -2.0, -1.0, 1.0, 1.0, 2.0, 2.0, 3.0, 4.0]
-
-    result = run_op_node([input_tensor], ov.round, "HALF_AWAY_FROM_ZERO")
-    assert np.allclose(result, expected)
+    assert node.get_type_name() == "Round"
+    assert node.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(node.get_output_shape(0)) == [9]
 
 
 def test_hsigmoid():
@@ -190,102 +168,42 @@ def test_hsigmoid():
 
 
 def test_gelu_operator_with_parameters():
-    runtime = get_runtime()
-
-    data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
-
     data_shape = [2, 2]
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
     model = ov.gelu(parameter_data, "erf")
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data_value)
-    expected = np.array([[-1.6391277e-06, 8.4134471e-01], [-4.5500278e-02, 2.9959502]], dtype=np.float32)
-    assert np.allclose(result, expected, 1e-6, 1e-6)
+    assert model.get_output_size() == 1
+    assert model.get_type_name() == "Gelu"
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [2, 2]
 
 
 def test_gelu_operator_with_array():
-    runtime = get_runtime()
-
     data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
 
     model = ov.gelu(data_value, "erf")
-    computation = runtime.computation(model)
-
-    result = computation()
-    expected = np.array([[-1.6391277e-06, 8.4134471e-01], [-4.5500278e-02, 2.9959502]], dtype=np.float32)
-    assert np.allclose(result, expected, 1e-6, 1e-6)
+    assert model.get_output_size() == 1
+    assert model.get_type_name() == "Gelu"
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [2, 2]
 
 
 def test_gelu_tanh_operator_with_parameters():
-    runtime = get_runtime()
-
-    data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
-
     data_shape = [2, 2]
     parameter_data = ov.parameter(data_shape, name="Data", dtype=np.float32)
 
     model = ov.gelu(parameter_data, "tanh")
-    computation = runtime.computation(model, parameter_data)
-
-    result = computation(data_value)
-    expected = np.array([[0.0, 0.841192], [-0.04540223, 2.9963627]], dtype=np.float32)
-    assert np.allclose(result, expected, 1e-6, 1e-6)
+    assert model.get_output_size() == 1
+    assert model.get_type_name() == "Gelu"
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [2, 2]
 
 
 def test_gelu_tanh_operator_with_array():
-    runtime = get_runtime()
-
     data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
 
     model = ov.gelu(data_value, "tanh")
-    computation = runtime.computation(model)
-
-    result = computation()
-    expected = np.array([[0.0, 0.841192], [-0.04540223, 2.9963627]], dtype=np.float32)
-
-    assert np.allclose(result, expected, 1e-6, 1e-6)
-
-
-type_tolerance = [
-    (np.float64, 1e-6),
-    (np.float32, 1e-6),
-    (np.float16, 1e-3),
-]
-
-
-@pytest.mark.parametrize("type_tolerance", type_tolerance)
-def test_softsign_with_parameters(type_tolerance):
-    dtype, atol = type_tolerance
-    data = np.random.uniform(-1.0, 1.0, (32, 5)).astype(dtype)
-    expected = np.divide(data, np.abs(data) + 1)
-
-    runtime = get_runtime()
-    param = ov.parameter(data.shape, dtype, name="Data")
-    result = runtime.computation(ov.softsign(param, "SoftSign"), param)(data)
-
-    assert np.allclose(result, expected, R_TOLERANCE, atol)
-
-
-@pytest.mark.parametrize("type_tolerance", type_tolerance)
-def test_softsign_with_array(type_tolerance):
-    dtype, atol = type_tolerance
-    data = np.random.uniform(-1.0, 1.0, (32, 5)).astype(dtype)
-    expected = np.divide(data, np.abs(data) + 1)
-
-    runtime = get_runtime()
-    result = runtime.computation(ov.softsign(data, "SoftSign"))()
-
-    assert np.allclose(result, expected, R_TOLERANCE, atol)
-
-
-@pytest.mark.parametrize("type_tolerance", type_tolerance)
-def test_softsign(type_tolerance):
-    dtype, atol = type_tolerance
-    data = np.random.uniform(-1.0, 1.0, (32, 5)).astype(dtype)
-    expected = np.divide(data, np.abs(data) + 1)
-
-    result = run_op_node([data], ov.softsign)
-
-    assert np.allclose(result, expected, R_TOLERANCE, atol)
+    assert model.get_output_size() == 1
+    assert model.get_type_name() == "Gelu"
+    assert model.get_output_element_type(0) == ov_runtime.Type.f32
+    assert list(model.get_output_shape(0)) == [2, 2]
diff --git a/src/bindings/python/tests/test_graph/test_pooling.py b/src/bindings/python/tests/test_graph/test_pooling.py
index 6ad24c43fee..91a5c0ea5d7 100644
--- a/src/bindings/python/tests/test_graph/test_pooling.py
+++ b/src/bindings/python/tests/test_graph/test_pooling.py
@@ -6,7 +6,7 @@ import numpy as np
 import pytest
 
 import openvino.runtime.opset8 as ov
-from tests.runtime import get_runtime
+from openvino.runtime import Type
 
 
 @pytest.fixture()
@@ -15,7 +15,6 @@ def ndarray_1x1x4x4():
 
 
 def test_avg_pool_2d(ndarray_1x1x4x4):
-    runtime = get_runtime()
     input_data = ndarray_1x1x4x4
     param = ov.parameter(input_data.shape, name="A", dtype=np.float32)
 
@@ -25,41 +24,15 @@ def test_avg_pool_2d(ndarray_1x1x4x4):
     pads_end = [0] * spatial_dim_count
     strides = [2, 2]
     exclude_pad = True
-    expected = [[[[13.5, 15.5], [21.5, 23.5]]]]
 
-    avg_pool_node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
-    computation = runtime.computation(avg_pool_node, param)
-    result = computation(input_data)
-    assert np.allclose(result, expected)
-
-    expected = [[[[13.5, 14.5, 15.5], [17.5, 18.5, 19.5], [21.5, 22.5, 23.5]]]]
-    strides = [1, 1]
-    avg_pool_node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
-    computation = runtime.computation(avg_pool_node, param)
-    result = computation(input_data)
-    assert np.allclose(result, expected)
-
-    pads_begin = [1, 1]
-    pads_end = [1, 1]
-    strides = [2, 2]
-    exclude_pad = True
-
-    expected = [[[[11.0, 12.5, 14.0], [17.0, 18.5, 20.0], [23.0, 24.5, 26.0]]]]
-    avg_pool_node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
-    computation = runtime.computation(avg_pool_node, param)
-    result = computation(input_data)
-    assert np.allclose(result, expected)
-
-    exclude_pad = False
-    expected = [[[[2.75, 6.25, 3.5], [8.5, 18.5, 10.0], [5.75, 12.25, 6.5]]]]
-    avg_pool_node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
-    computation = runtime.computation(avg_pool_node, param)
-    result = computation(input_data)
-    assert np.allclose(result, expected)
+    node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
+    assert node.get_type_name() == "AvgPool"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 1, 2, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_avg_pooling_3d(ndarray_1x1x4x4):
-    rt = get_runtime()
     data = ndarray_1x1x4x4
     data = np.broadcast_to(data, (1, 1, 4, 4, 4))
     param = ov.parameter(list(data.shape))
@@ -70,21 +43,14 @@ def test_avg_pooling_3d(ndarray_1x1x4x4):
     pads_end = [0] * spatial_dim_count
     exclude_pad = True
 
-    avgpool = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
-    comp = rt.computation(avgpool, param)
-    result = comp(data)
-    result_ref = [[[[[13.5, 15.5], [21.5, 23.5]], [[13.5, 15.5], [21.5, 23.5]]]]]
-    assert np.allclose(result, result_ref)
+    node = ov.avg_pool(param, strides, pads_begin, pads_end, kernel_shape, exclude_pad)
+    assert node.get_type_name() == "AvgPool"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 1, 2, 2, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_max_pool_basic():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
@@ -96,7 +62,7 @@ def test_max_pool_basic():
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -107,23 +73,15 @@ def test_max_pool_basic():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array([[[[5.5, 6.5, 7.5], [9.5, 10.5, 11.5], [13.5, 14.5, 15.5]]]], dtype=np.float32)
-    expected_idx = np.array([[[[5, 6, 7], [9, 10, 11], [13, 14, 15]]]], dtype=np.int32)
-    assert np.allclose(result[0], expected)
-    assert np.allclose(result[1], expected_idx)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 3, 3]
+    assert list(node.get_output_shape(1)) == [1, 1, 3, 3]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_strides():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [2, 1]
     dilations = [1, 1]
@@ -135,7 +93,7 @@ def test_max_pool_strides():
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -146,23 +104,15 @@ def test_max_pool_strides():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array([[[[5.5, 6.5, 7.5], [13.5, 14.5, 15.5]]]], dtype=np.float32)
-    expected_idx = np.array([[[[5, 6, 7], [13, 14, 15]]]], dtype=np.int32)
-    assert np.allclose(result[0], expected)
-    assert np.allclose(result[1], expected_idx)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 2, 3]
+    assert list(node.get_output_shape(1)) == [1, 1, 2, 3]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_kernel_shape1x1():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
@@ -174,7 +124,7 @@ def test_max_pool_kernel_shape1x1():
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -185,21 +135,15 @@ def test_max_pool_kernel_shape1x1():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    assert np.allclose(result[0], data)
-    assert np.allclose(result[1], np.arange(0, 16, dtype=np.int32).reshape((1, 1, 4, 4)))
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 4, 4]
+    assert list(node.get_output_shape(1)) == [1, 1, 4, 4]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_kernel_shape3x3():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
@@ -211,7 +155,7 @@ def test_max_pool_kernel_shape3x3():
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -222,40 +166,27 @@ def test_max_pool_kernel_shape3x3():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array([[[[10.5, 11.5], [14.5, 15.5]]]], dtype=np.float32)
-    assert np.allclose(result[0], expected)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 2, 2]
+    assert list(node.get_output_shape(1)) == [1, 1, 2, 2]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_non_zero_pads():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
     pads_begin = [1, 1]
     pads_end = [1, 1]
-    """0   0  ,  0  ,  0  ,  0,    0
-       0 [ 0.5,  1.5,  2.5,  3.5], 0,
-       0 [ 4.5,  5.5,  6.5,  7.5], 0,
-       0 [ 8.5,  9.5, 10.5, 11.5], 0,
-       0 [12.5, 13.5, 14.5, 15.5], 0
-       0   0  ,  0  ,  0  ,  0,    0
-    """
     kernel_shape = [2, 2]
     rounding_type = "floor"
     auto_pad = None
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -266,67 +197,27 @@ def test_max_pool_non_zero_pads():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array(
-        [
-            [
-                [
-                    [0.5, 1.5, 2.5, 3.5, 3.5],
-                    [4.5, 5.5, 6.5, 7.5, 7.5],
-                    [8.5, 9.5, 10.5, 11.5, 11.5],
-                    [12.5, 13.5, 14.5, 15.5, 15.5],
-                    [12.5, 13.5, 14.5, 15.5, 15.5],
-                ],
-            ],
-        ],
-        dtype=np.float32,
-    )
-    expected_idx = np.array(
-        [
-            [
-                [
-                    [0, 1, 2, 3, 3],
-                    [4, 5, 6, 7, 7],
-                    [8, 9, 10, 11, 11],
-                    [12, 13, 14, 15, 15],
-                    [12, 13, 14, 15, 15],
-                ],
-            ],
-        ],
-        dtype=np.int32,
-    )
-    assert np.allclose(result[0], expected)
-    assert np.allclose(result[1], expected_idx)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 5, 5]
+    assert list(node.get_output_shape(1)) == [1, 1, 5, 5]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_same_upper_auto_pads():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
     pads_begin = [0, 0]
     pads_end = [0, 0]
-    """[ 0.5,  1.5,  2.5,  3.5], 0,
-       [ 4.5,  5.5,  6.5,  7.5], 0,
-       [ 8.5,  9.5, 10.5, 11.5], 0,
-       [12.5, 13.5, 14.5, 15.5], 0
-         0  ,  0  ,  0  ,  0,    0
-    """
     kernel_shape = [2, 2]
     auto_pad = "same_upper"
     rounding_type = "floor"
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -337,65 +228,27 @@ def test_max_pool_same_upper_auto_pads():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array(
-        [
-            [
-                [
-                    [5.5, 6.5, 7.5, 7.5],
-                    [9.5, 10.5, 11.5, 11.5],
-                    [13.5, 14.5, 15.5, 15.5],
-                    [13.5, 14.5, 15.5, 15.5],
-                ],
-            ],
-        ],
-        dtype=np.float32,
-    )
-    expected_idx = np.array(
-        [
-            [
-                [
-                    [5, 6, 7, 7],
-                    [9, 10, 11, 11],
-                    [13, 14, 15, 15],
-                    [13, 14, 15, 15],
-                ],
-            ],
-        ],
-        dtype=np.int32,
-    )
-    assert np.allclose(result[0], expected)
-    assert np.allclose(result[1], expected_idx)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 4, 4]
+    assert list(node.get_output_shape(1)) == [1, 1, 4, 4]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
 
 
 def test_max_pool_same_lower_auto_pads():
-    rt = get_runtime()
-
-    """array([[[[ 0.5,  1.5,  2.5,  3.5],
-               [ 4.5,  5.5,  6.5,  7.5],
-               [ 8.5,  9.5, 10.5, 11.5],
-               [12.5, 13.5, 14.5, 15.5]]]], dtype=float32)
-    """
     data = np.arange(0.5, 16, dtype=np.float32).reshape((1, 1, 4, 4))
     strides = [1, 1]
     dilations = [1, 1]
     pads_begin = [0, 0]
     pads_end = [0, 0]
-    """0   0  ,  0  ,  0  ,  0,
-       0 [ 0.5,  1.5,  2.5,  3.5],
-       0 [ 4.5,  5.5,  6.5,  7.5],
-       0 [ 8.5,  9.5, 10.5, 11.5],
-       0 [12.5, 13.5, 14.5, 15.5],
-    """
     kernel_shape = [2, 2]
     auto_pad = "same_lower"
     rounding_type = "floor"
     index_et = "i32"
 
     data_node = ov.parameter(data.shape, name="A", dtype=np.float32)
-    maxpool_node = ov.max_pool(
+    node = ov.max_pool(
         data_node,
         strides,
         dilations,
@@ -406,34 +259,9 @@ def test_max_pool_same_lower_auto_pads():
         auto_pad,
         index_et,
     )
-    comp = rt.computation(maxpool_node, data_node)
-    result = comp(data)
-
-    expected = np.array(
-        [
-            [
-                [
-                    [0.5, 1.5, 2.5, 3.5],
-                    [4.5, 5.5, 6.5, 7.5],
-                    [8.5, 9.5, 10.5, 11.5],
-                    [12.5, 13.5, 14.5, 15.5],
-                ],
-            ],
-        ],
-        dtype=np.float32,
-    )
-    expected_idx = np.array(
-        [
-            [
-                [
-                    [0, 1, 2, 3],
-                    [4, 5, 6, 7],
-                    [8, 9, 10, 11],
-                    [12, 13, 14, 15],
-                ],
-            ],
-        ],
-        dtype=np.int32,
-    )
-    assert np.allclose(result[0], expected)
-    assert np.allclose(result[1], expected_idx)
+    assert node.get_type_name() == "MaxPool"
+    assert node.get_output_size() == 2
+    assert list(node.get_output_shape(0)) == [1, 1, 4, 4]
+    assert list(node.get_output_shape(1)) == [1, 1, 4, 4]
+    assert node.get_output_element_type(0) == Type.f32
+    assert node.get_output_element_type(1) == Type.i32
diff --git a/src/bindings/python/tests/test_graph/test_preprocess.py b/src/bindings/python/tests/test_graph/test_preprocess.py
index 411903b821e..247763fd966 100644
--- a/src/bindings/python/tests/test_graph/test_preprocess.py
+++ b/src/bindings/python/tests/test_graph/test_preprocess.py
@@ -10,7 +10,6 @@ import openvino.runtime.opset8 as ops
 from openvino.runtime import Model, Output, Type
 from openvino.runtime.utils.decorators import custom_preprocess_function
 from openvino.runtime import Core
-from tests.runtime import get_runtime
 from openvino.preprocess import PrePostProcessor, ColorFormat, ResizeAlgorithm
 
 
@@ -19,20 +18,18 @@ def test_graph_preprocess_mean():
     parameter_a = ops.parameter(shape, dtype=np.float32, name="A")
     model = parameter_a
     function = Model(model, [parameter_a], "TestFunction")
-
     ppp = PrePostProcessor(function)
     inp = ppp.input()
     prep = inp.preprocess()
     prep.mean(1.0)
     function = ppp.build()
-
-    input_data = np.array([[1, 2], [3, 4]]).astype(np.float32)
-    expected_output = np.array([[0, 1], [2, 3]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ordered_ops()]
+    assert len(model_operators) == 4
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 2]
+    assert function.get_output_element_type(0) == Type.f32
+    assert "Constant" in model_operators
+    assert "Subtract" in model_operators
 
 
 def test_graph_preprocess_mean_vector():
@@ -47,13 +44,13 @@ def test_graph_preprocess_mean_vector():
     ppp.input().preprocess().mean([1., 2.])
     function = ppp.build()
 
-    input_data = np.array([[1, 2], [3, 4]]).astype(np.float32)
-    expected_output = np.array([[0, 0], [2, 2]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ordered_ops()]
+    assert len(model_operators) == 4
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 2]
+    assert function.get_output_element_type(0) == Type.f32
+    assert "Constant" in model_operators
+    assert "Subtract" in model_operators
 
 
 def test_graph_preprocess_scale_vector():
@@ -69,13 +66,13 @@ def test_graph_preprocess_scale_vector():
     inp.preprocess().scale([0.5, 2.0])
     function = ppp.build()
 
-    input_data = np.array([[1, 2], [3, 4]]).astype(np.float32)
-    expected_output = np.array([[2, 1], [6, 2]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ordered_ops()]
+    assert len(model_operators) == 4
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 2]
+    assert function.get_output_element_type(0) == Type.f32
+    assert "Constant" in model_operators
+    assert "Divide" in model_operators
 
 
 def test_graph_preprocess_mean_scale_convert():
@@ -97,16 +94,24 @@ def test_graph_preprocess_mean_scale_convert():
     inp1.preprocess().convert_element_type(Type.f32).mean(1.).custom(custom_preprocess)
     function = ppp.build()
 
-    input_data1 = np.array([[0, 1], [2, -2]]).astype(np.int32)
-    input_data2 = np.array([[1, 3], [5, 7]]).astype(np.int32)
-    expected_output1 = np.array([[1, 0], [1, 3]]).astype(np.float32)
-    expected_output2 = np.array([[0, 1], [2, 3]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    [output1, output2] = computation(input_data1, input_data2)
-    assert np.equal(output1, expected_output1).all()
-    assert np.equal(output2, expected_output2).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Convert",
+        "Constant",
+        "Subtract",
+        "Divide",
+        "Result",
+        "Abs",
+    ]
+    assert len(model_operators) == 15
+    assert function.get_output_size() == 2
+    assert list(function.get_output_shape(0)) == [2, 2]
+    assert list(function.get_output_shape(1)) == [2, 2]
+    assert function.get_output_element_type(0) == Type.i32
+    assert function.get_output_element_type(1) == Type.i32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_input_output_by_name():
@@ -131,16 +136,24 @@ def test_graph_preprocess_input_output_by_name():
     out2.postprocess().custom(custom_preprocess)
     function = ppp.build()
 
-    input_data1 = np.array([[0, 1], [2, -2]]).astype(np.int32)
-    input_data2 = np.array([[-1, 3], [5, 7]]).astype(np.int32)
-    expected_output1 = np.array([[1, 0], [1, 3]]).astype(np.float32)
-    expected_output2 = np.array([[1, 1], [2, 3]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    [output1, output2] = computation(input_data1, input_data2)
-    assert np.equal(output1, expected_output1).all()
-    assert np.equal(output2, expected_output2).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Convert",
+        "Constant",
+        "Subtract",
+        "Divide",
+        "Result",
+        "Abs",
+    ]
+    assert len(model_operators) == 16
+    assert function.get_output_size() == 2
+    assert list(function.get_output_shape(0)) == [2, 2]
+    assert list(function.get_output_shape(1)) == [2, 2]
+    assert function.get_output_element_type(0) == Type.i32
+    assert function.get_output_element_type(1) == Type.i32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_output_postprocess():
@@ -155,7 +168,6 @@ def test_graph_preprocess_output_postprocess():
     @custom_preprocess_function
     def custom_postprocess(output: Output):
         return ops.abs(output)
-
     ppp = PrePostProcessor(function)
     inp = ppp.input()
     inp.tensor().set_layout(layout1)
@@ -168,13 +180,22 @@ def test_graph_preprocess_output_postprocess():
     out.postprocess().custom(custom_postprocess).convert_element_type(Type.f16).convert_element_type()
     function = ppp.build()
 
-    input_data = np.array([[-1, -2, -3], [-4, -5, -6]]).astype(np.int32)
-    expected_output = np.array([[2, 4, 6], [5, 7, 9]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Convert",
+        "Constant",
+        "Subtract",
+        "Transpose",
+        "Result",
+        "Abs",
+    ]
+    assert len(model_operators) == 14
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 3]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_spatial_static_shape():
@@ -196,13 +217,20 @@ def test_graph_preprocess_spatial_static_shape():
     out.model().set_layout(layout)
     function = ppp.build()
 
-    input_data = np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]]).astype(np.int32)
-    expected_output = np.array([[[0, 1], [2, 3]], [[3, 4], [5, 6]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Convert",
+        "Constant",
+        "Subtract",
+        "Result",
+    ]
+    assert len(model_operators) == 7
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 2, 2]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_set_shape():
@@ -225,15 +253,19 @@ def test_graph_preprocess_set_shape():
     inp.preprocess().custom(custom_crop)
     function = ppp.build()
 
-    input_data = np.array([[[0, 1, 2], [3, 4, 5], [6, 7, 8]],
-                           [[9, 10, 11], [12, 13, 14], [15, 16, 17]],
-                           [[18, 19, 20], [21, 22, 23], [24, 25, 26]]]).astype(np.int32)
-    expected_output = np.array([[[13]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Slice",
+    ]
+    assert len(model_operators) == 7
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 1, 1]
+    assert function.get_output_element_type(0) == Type.i32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_set_from_tensor():
@@ -282,12 +314,20 @@ def test_graph_preprocess_set_from_np_infer():
     assert function.input().shape == ov.Shape([3, 3, 3])
     assert function.input().element_type == Type.i32
 
-    expected_output = np.array([[[13]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Convert",
+        "Constant",
+        "Result",
+        "Slice",
+    ]
+    assert len(model_operators) == 8
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 1, 1]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_set_memory_type():
@@ -344,13 +384,20 @@ def test_graph_preprocess_steps(algorithm, color_format1, color_format2, is_fail
         assert "is not convertible to" in str(e.value)
     else:
         function = custom_processor.build()
-        input_data = np.array([[[[1, 2, 3], [4, 5, 6], [7, 8, 9]]]]).astype(np.float32)
-        expected_output = np.array([[[[0, 3, 6], [1, 4, 7], [2, 5, 8]]]]).astype(np.float32)
-
-        runtime = get_runtime()
-        computation = runtime.computation(function)
-        output = computation(input_data)
-        assert np.equal(output, expected_output).all()
+        model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+        expected_ops = [
+            "Parameter",
+            "Constant",
+            "Result",
+            "Gather",
+            "Interpolate",
+        ]
+        assert len(model_operators) == 16
+        assert function.get_output_size() == 1
+        assert list(function.get_output_shape(0)) == [1, 1, 3, 3]
+        assert function.get_output_element_type(0) == Type.f32
+        for op in expected_ops:
+            assert op in model_operators
 
 
 def test_graph_preprocess_postprocess_layout():
@@ -369,13 +416,21 @@ def test_graph_preprocess_postprocess_layout():
     out.postprocess().convert_layout([0, 1, 2, 3])
     function = ppp.build()
 
-    input_data = np.array([[[[1, 2, 3], [4, 5, 6], [7, 8, 9]]]]).astype(np.float32)
-    expected_output = np.array([[[[0, 3, 6], [1, 4, 7], [2, 5, 8]]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Gather",
+        "Range",
+        "Transpose",
+    ]
+    assert len(model_operators) == 14
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 1, 3, 3]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_reverse_channels():
@@ -391,13 +446,20 @@ def test_graph_preprocess_reverse_channels():
     inp.preprocess().mean(1.).reverse_channels()
     function = ppp.build()
 
-    input_data = np.array([[[[1, 2], [3, 4]], [[5, 6], [7, 8]]]]).astype(np.float32)
-    expected_output = np.array([[[[4, 5], [6, 7]], [[0, 1], [2, 3]]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Gather",
+        "Range",
+    ]
+    assert len(model_operators) == 10
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 2, 2, 2]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_crop():
@@ -412,13 +474,20 @@ def test_graph_preprocess_crop():
     ppp.input().preprocess().crop([0, 0, 1, 1], [1, 2, -1, -1])
     function = ppp.build()
 
-    input_data = np.arange(18).astype(np.float32).reshape(tensor_shape)
-    expected_output = np.array([4, 13]).astype(np.float32).reshape(orig_shape)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Relu",
+        "Slice",
+    ]
+    assert len(model_operators) == 7
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 2, 1, 1]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_resize_algorithm():
@@ -435,13 +504,20 @@ def test_graph_preprocess_resize_algorithm():
     inp.preprocess().mean(1.).resize(resize_alg, 3, 3)
     function = ppp.build()
 
-    input_data = np.array([[[[1, 2, 3], [4, 5, 6], [7, 8, 9]]]]).astype(np.float32)
-    expected_output = np.array([[[[0, 1, 2], [3, 4, 5], [6, 7, 8]]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data)
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Subtract",
+        "Interpolate",
+    ]
+    assert len(model_operators) == 8
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [1, 1, 3, 3]
+    assert function.get_output_element_type(0) == Type.f32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_model():
@@ -517,14 +593,23 @@ def test_graph_preprocess_model():
     ppp.output(0).postprocess().custom(custom_preprocess)
     function = ppp.build()
 
-    input_data = np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]]).astype(np.float32)
-    expected_output = np.array([[[2, 1], [4, 7]], [[10, 13], [16, 19]]]).astype(np.float32)
-
-    runtime = get_runtime()
-    computation = runtime.computation(function)
-    output = computation(input_data, input_data)
-
-    assert np.equal(output, expected_output).all()
+    model_operators = [op.get_name().split("_")[0] for op in function.get_ops()]
+    expected_ops = [
+        "Parameter",
+        "Constant",
+        "Result",
+        "Subtract",
+        "Convert",
+        "Abs",
+        "Add",
+        "Divide",
+    ]
+    assert len(model_operators) == 13
+    assert function.get_output_size() == 1
+    assert list(function.get_output_shape(0)) == [2, 2, 2]
+    assert function.get_output_element_type(0) == Type.i32
+    for op in expected_ops:
+        assert op in model_operators
 
 
 def test_graph_preprocess_dump():
diff --git a/src/bindings/python/tests/test_graph/test_random_uniform.py b/src/bindings/python/tests/test_graph/test_random_uniform.py
index fac2f66b608..dcc42fb2930 100644
--- a/src/bindings/python/tests/test_graph/test_random_uniform.py
+++ b/src/bindings/python/tests/test_graph/test_random_uniform.py
@@ -2,30 +2,20 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
-import openvino.runtime.opset8 as ov
+import openvino.runtime as ov
+import openvino.runtime.opset8 as ops
 import numpy as np
-from tests.runtime import get_runtime
 
 
 def test_random_uniform():
-    runtime = get_runtime()
-    input_tensor = ov.constant(np.array([2, 4, 3], dtype=np.int32))
-    min_val = ov.constant(np.array([-2.7], dtype=np.float32))
-    max_val = ov.constant(np.array([3.5], dtype=np.float32))
+    input_tensor = ops.constant(np.array([2, 4, 3], dtype=np.int32))
+    min_val = ops.constant(np.array([-2.7], dtype=np.float32))
+    max_val = ops.constant(np.array([3.5], dtype=np.float32))
 
-    random_uniform_node = ov.random_uniform(input_tensor, min_val, max_val,
-                                            output_type="f32", global_seed=7461,
-                                            op_seed=1546)
-    computation = runtime.computation(random_uniform_node)
-    random_uniform_results = computation()
-    expected_results = np.array([[[2.8450181, -2.3457108, 2.2134445],
-                                  [-1.0436587, 0.79548645, 1.3023183],
-                                  [0.34447956, -2.0267959, 1.3989122],
-                                  [0.9607613, 1.5363653, 3.117298]],
-
-                                 [[1.570041, 2.2782724, 2.3193843],
-                                  [3.3393657, 0.63299894, 0.41231918],
-                                  [3.1739233, 0.03919673, -0.2136085],
-                                  [-1.4519991, -2.277353, 2.630727]]], dtype=np.float32)
-
-    assert np.allclose(random_uniform_results, expected_results)
+    random_uniform_node = ops.random_uniform(input_tensor, min_val, max_val,
+                                             output_type="f32", global_seed=7461,
+                                             op_seed=1546)
+    assert random_uniform_node.get_output_size() == 1
+    assert random_uniform_node.get_type_name() == "RandomUniform"
+    assert random_uniform_node.get_output_element_type(0) == ov.Type.f32
+    assert list(random_uniform_node.get_output_shape(0)) == [2, 4, 3]
diff --git a/src/bindings/python/tests/test_graph/test_rdft.py b/src/bindings/python/tests/test_graph/test_rdft.py
index 3c3392f72aa..849fd7c6a29 100644
--- a/src/bindings/python/tests/test_graph/test_rdft.py
+++ b/src/bindings/python/tests/test_graph/test_rdft.py
@@ -3,163 +3,137 @@
 # SPDX-License-Identifier: Apache-2.0
 
 import openvino.runtime.opset9 as ov
-from openvino.runtime import Shape
+from openvino.runtime import Shape, Type
 import numpy as np
-from tests.runtime import get_runtime
 
 
 np.random.seed(0)
 
 
 def test_rdft_1d():
-    runtime = get_runtime()
     input_size = 50
     shape = [input_size]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     input_axes = ov.constant(np.array([0], dtype=np.int64))
 
     node = ov.rdft(param, input_axes)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    np_results = np.fft.rfft(data)
-    expected_results = np.stack((np_results.real, np_results.imag), axis=-1)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "RDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [26, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_irdft_1d():
-    runtime = get_runtime()
     signal_size = 50
     shape = [signal_size // 2 + 1, 2]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     input_axes = ov.constant(np.array([0], dtype=np.int64))
     node = ov.irdft(param, input_axes, ov.constant(np.array([signal_size], dtype=np.int64)))
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    expected_results = np.fft.irfft(data[:, 0] + 1j * data[:, 1], signal_size)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "IRDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [50]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_rdft_2d():
-    runtime = get_runtime()
     shape = [100, 128]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [0, 1]
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     node = ov.rdft(param, input_axes)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    np_results = np.fft.rfftn(data, axes=axes)
-    expected_results = np.stack((np_results.real, np_results.imag), axis=-1)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0007)
+    assert node.get_type_name() == "RDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [100, 65, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_rdft_2d_signal_size():
-    runtime = get_runtime()
     shape = [100, 128]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [0, 1]
     signal_size = [30, 40]
     axes_node = ov.constant(np.array(axes, dtype=np.int64))
     signal_size_node = ov.constant(np.array(signal_size, dtype=np.int64))
     node = ov.rdft(param, axes_node, signal_size_node)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    np_results = np.fft.rfftn(data, s=signal_size, axes=axes)
-    expected_results = np.stack((np_results.real, np_results.imag), axis=-1)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0007)
+    assert node.get_type_name() == "RDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [30, 21, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_irdft_2d():
-    runtime = get_runtime()
     axes = [0, 1]
     input_shape = [100, 65, 2]
-    data = np.random.uniform(0, 1, input_shape).astype(np.float32)
     param = ov.parameter(Shape(input_shape), name="input", dtype=np.float32)
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     node = ov.irdft(param, input_axes)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    expected_results = np.fft.irfftn(data[:, :, 0] + 1j * data[:, :, 1], axes=axes)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "IRDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [100, 128]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_irdft_2d_signal_size():
-    runtime = get_runtime()
     axes = [0, 1]
     input_shape = [100, 65, 2]
     signal_size = [100, 65]
-    data = np.random.uniform(0, 1, input_shape).astype(np.float32)
     param = ov.parameter(Shape(input_shape), name="input", dtype=np.float32)
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     signal_size_node = ov.constant(np.array(signal_size, dtype=np.int64))
     node = ov.irdft(param, input_axes, signal_size_node)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    expected_results = np.fft.irfftn(data[:, :, 0] + 1j * data[:, :, 1], s=signal_size, axes=axes)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "IRDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [100, 65]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_rdft_4d():
-    runtime = get_runtime()
     shape = [1, 192, 36, 64]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [-2, -1]
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     node = ov.rdft(param, input_axes)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    np_results = np.fft.rfftn(data, axes=axes)
-    expected_results = np.stack((np_results.real, np_results.imag), axis=-1)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0007)
+    assert node.get_type_name() == "RDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 192, 36, 33, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_rdft_4d_signal_size():
-    runtime = get_runtime()
     shape = [1, 192, 36, 64]
     signal_size = [36, 64]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [-2, -1]
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     signal_size_node = ov.constant(np.array(signal_size, dtype=np.int64))
     node = ov.rdft(param, input_axes, signal_size_node)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    np_results = np.fft.rfftn(data, signal_size, axes=axes)
-    expected_results = np.stack((np_results.real, np_results.imag), axis=-1)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0007)
+    assert node.get_type_name() == "RDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 192, 36, 33, 2]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_irdft_4d():
-    runtime = get_runtime()
     shape = [1, 192, 36, 33, 2]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [-2, -1]
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     node = ov.irdft(param, input_axes)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    expected_results = np.fft.irfftn(data[:, :, :, :, 0] + 1j * data[:, :, :, :, 1], axes=axes)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "IRDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 192, 36, 64]
+    assert node.get_output_element_type(0) == Type.f32
 
 
 def test_irdft_4d_signal_size():
-    runtime = get_runtime()
     shape = [1, 192, 36, 33, 2]
     signal_size = [36, 64]
-    data = np.random.uniform(0, 1, shape).astype(np.float32)
     param = ov.parameter(Shape(shape), name="input", dtype=np.float32)
     axes = [-2, -1]
     input_axes = ov.constant(np.array(axes, dtype=np.int64))
     signal_size_node = ov.constant(np.array(signal_size, dtype=np.int64))
     node = ov.irdft(param, input_axes, signal_size_node)
-    computation = runtime.computation(node, param)
-    actual = computation(data)
-    expected_results = np.fft.irfftn(data[:, :, :, :, 0] + 1j * data[:, :, :, :, 1], signal_size, axes=axes)
-    np.testing.assert_allclose(expected_results, actual[0], atol=0.0001)
+    assert node.get_type_name() == "IRDFT"
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == [1, 192, 36, 64]
+    assert node.get_output_element_type(0) == Type.f32
diff --git a/src/bindings/python/tests/test_graph/test_reduction.py b/src/bindings/python/tests/test_graph/test_reduction.py
index 8adaf02b914..adcdfc08418 100644
--- a/src/bindings/python/tests/test_graph/test_reduction.py
+++ b/src/bindings/python/tests/test_graph/test_reduction.py
@@ -4,60 +4,56 @@
 
 import numpy as np
 import pytest
-from openvino.runtime import PartialShape, Dimension
 
 import openvino.runtime.opset9 as ov
-from openvino.runtime.utils.types import make_constant_node
-from tests.runtime import get_runtime
-from tests.test_graph.util import run_op_node
 
 
 @pytest.mark.parametrize(
-    ("graph_api_helper", "numpy_function", "reduction_axes"),
+    ("graph_api_helper", "reduction_axes", "expected_shape"),
     [
-        (ov.reduce_max, np.max, np.array([0, 1, 2, 3])),
-        (ov.reduce_min, np.min, np.array([0, 1, 2, 3])),
-        (ov.reduce_sum, np.sum, np.array([0, 1, 2, 3])),
-        (ov.reduce_prod, np.prod, np.array([0, 1, 2, 3])),
-        (ov.reduce_max, np.max, np.array([0])),
-        (ov.reduce_min, np.min, np.array([0])),
-        (ov.reduce_sum, np.sum, np.array([0])),
-        (ov.reduce_prod, np.prod, np.array([0])),
-        (ov.reduce_max, np.max, np.array([0, 2])),
-        (ov.reduce_min, np.min, np.array([0, 2])),
-        (ov.reduce_sum, np.sum, np.array([0, 2])),
-        (ov.reduce_prod, np.prod, np.array([0, 2])),
+        (ov.reduce_max, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_min, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_sum, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_prod, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_max, np.array([0]), [4, 3, 2]),
+        (ov.reduce_min, np.array([0]), [4, 3, 2]),
+        (ov.reduce_sum, np.array([0]), [4, 3, 2]),
+        (ov.reduce_prod, np.array([0]), [4, 3, 2]),
+        (ov.reduce_max, np.array([0, 2]), [4, 2]),
+        (ov.reduce_min, np.array([0, 2]), [4, 2]),
+        (ov.reduce_sum, np.array([0, 2]), [4, 2]),
+        (ov.reduce_prod, np.array([0, 2]), [4, 2]),
     ],
 )
-def test_reduction_ops(graph_api_helper, numpy_function, reduction_axes):
+def test_reduction_ops(graph_api_helper, reduction_axes, expected_shape):
     shape = [2, 4, 3, 2]
     np.random.seed(133391)
     input_data = np.random.randn(*shape).astype(np.float32)
 
-    expected = numpy_function(input_data, axis=tuple(reduction_axes))
-    result = run_op_node([input_data], graph_api_helper, reduction_axes)
-    assert np.allclose(result, expected)
+    node = graph_api_helper(input_data, reduction_axes)
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == expected_shape
 
 
 @pytest.mark.parametrize(
-    ("graph_api_helper", "numpy_function", "reduction_axes"),
+    ("graph_api_helper", "reduction_axes", "expected_shape"),
     [
-        (ov.reduce_logical_and, np.logical_and.reduce, np.array([0])),
-        (ov.reduce_logical_or, np.logical_or.reduce, np.array([0])),
-        (ov.reduce_logical_and, np.logical_and.reduce, np.array([0, 2])),
-        (ov.reduce_logical_or, np.logical_or.reduce, np.array([0, 2])),
-        (ov.reduce_logical_and, np.logical_and.reduce, np.array([0, 1, 2, 3])),
-        (ov.reduce_logical_or, np.logical_or.reduce, np.array([0, 1, 2, 3])),
+        (ov.reduce_logical_and, np.array([0]), [4, 3, 2]),
+        (ov.reduce_logical_or, np.array([0]), [4, 3, 2]),
+        (ov.reduce_logical_and, np.array([0, 2]), [4, 2]),
+        (ov.reduce_logical_or, np.array([0, 2]), [4, 2]),
+        (ov.reduce_logical_and, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_logical_or, np.array([0, 1, 2, 3]), []),
     ],
 )
-def test_reduction_logical_ops(graph_api_helper, numpy_function, reduction_axes):
+def test_reduction_logical_ops(graph_api_helper, reduction_axes, expected_shape):
     shape = [2, 4, 3, 2]
     np.random.seed(133391)
     input_data = np.random.randn(*shape).astype(bool)
 
-    expected = numpy_function(input_data, axis=tuple(reduction_axes))
-    result = run_op_node([input_data], graph_api_helper, reduction_axes)
-    assert np.allclose(result, expected)
+    node = graph_api_helper(input_data, reduction_axes)
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == expected_shape
 
 
 def test_topk():
@@ -73,21 +69,21 @@ def test_topk():
 
 
 @pytest.mark.parametrize(
-    ("graph_api_helper", "numpy_function", "reduction_axes"),
+    ("graph_api_helper", "reduction_axes", "expected_shape"),
     [
-        (ov.reduce_mean, np.mean, np.array([0, 1, 2, 3])),
-        (ov.reduce_mean, np.mean, np.array([0])),
-        (ov.reduce_mean, np.mean, np.array([0, 2])),
+        (ov.reduce_mean, np.array([0, 1, 2, 3]), []),
+        (ov.reduce_mean, np.array([0]), [4, 3, 2]),
+        (ov.reduce_mean, np.array([0, 2]), [4, 2]),
     ],
 )
-def test_reduce_mean_op(graph_api_helper, numpy_function, reduction_axes):
+def test_reduce_mean_op(graph_api_helper, reduction_axes, expected_shape):
     shape = [2, 4, 3, 2]
     np.random.seed(133391)
     input_data = np.random.randn(*shape).astype(np.float32)
 
-    expected = numpy_function(input_data, axis=tuple(reduction_axes))
-    result = run_op_node([input_data], graph_api_helper, reduction_axes)
-    assert np.allclose(result, expected)
+    node = graph_api_helper(input_data, reduction_axes)
+    assert node.get_output_size() == 1
+    assert list(node.get_output_shape(0)) == expected_shape
 
 
 def test_non_zero():
@@ -141,16 +137,10 @@ def test_roi_align():
 def test_cum_sum(input_shape, cumsum_axis, reverse):
     input_data = np.arange(np.prod(input_shape)).reshape(input_shape)
 
-    if reverse:
-        expected = np.cumsum(input_data[::-1], axis=cumsum_axis)[::-1]
-    else:
-        expected = np.cumsum(input_data, axis=cumsum_axis)
-
-    runtime = get_runtime()
     node = ov.cum_sum(input_data, cumsum_axis, reverse=reverse)
-    computation = runtime.computation(node)
-    result = computation()
-    assert np.allclose(result, expected)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "CumSum"
+    assert list(node.get_output_shape(0)) == input_shape
 
 
 def test_normalize_l2():
@@ -161,38 +151,7 @@ def test_normalize_l2():
     eps = 1e-6
     eps_mode = "add"
 
-    runtime = get_runtime()
     node = ov.normalize_l2(input_data, axes, eps, eps_mode)
-    computation = runtime.computation(node)
-    result = computation()
-
-    expected = np.array(
-        [
-            0.01428571,
-            0.02857143,
-            0.04285714,
-            0.05714286,
-            0.07142857,
-            0.08571429,
-            0.1,
-            0.11428571,
-            0.12857144,
-            0.14285715,
-            0.15714286,
-            0.17142858,
-            0.18571429,
-            0.2,
-            0.21428572,
-            0.22857143,
-            0.24285714,
-            0.25714287,
-            0.27142859,
-            0.2857143,
-            0.30000001,
-            0.31428573,
-            0.32857144,
-            0.34285715,
-        ],
-    ).reshape(input_shape)
-
-    assert np.allclose(result, expected)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "NormalizeL2"
+    assert list(node.get_output_shape(0)) == input_shape
diff --git a/src/bindings/python/tests/test_graph/test_roll.py b/src/bindings/python/tests/test_graph/test_roll.py
index 7496abb7f28..c0e2fd99066 100644
--- a/src/bindings/python/tests/test_graph/test_roll.py
+++ b/src/bindings/python/tests/test_graph/test_roll.py
@@ -4,19 +4,15 @@
 
 import openvino.runtime.opset8 as ov
 import numpy as np
-from tests.runtime import get_runtime
 
 
 def test_roll():
-    runtime = get_runtime()
     input_vals = np.reshape(np.arange(10), (2, 5))
     input_tensor = ov.constant(input_vals)
     input_shift = ov.constant(np.array([-10, 7], dtype=np.int32))
     input_axes = ov.constant(np.array([-1, 0], dtype=np.int32))
 
     roll_node = ov.roll(input_tensor, input_shift, input_axes)
-    computation = runtime.computation(roll_node)
-    roll_results = computation()
-    expected_results = np.roll(input_vals, shift=(-10, 7), axis=(-1, 0))
-
-    assert np.allclose(roll_results, expected_results)
+    assert roll_node.get_output_size() == 1
+    assert roll_node.get_type_name() == "Roll"
+    assert list(roll_node.get_output_shape(0)) == [2, 5]
diff --git a/src/bindings/python/tests/test_graph/test_sequence_processing.py b/src/bindings/python/tests/test_graph/test_sequence_processing.py
index 3894739499b..f4242927d1a 100644
--- a/src/bindings/python/tests/test_graph/test_sequence_processing.py
+++ b/src/bindings/python/tests/test_graph/test_sequence_processing.py
@@ -5,20 +5,14 @@
 import numpy as np
 
 import openvino.runtime.opset8 as ov
-from tests.runtime import get_runtime
-from tests.test_graph.util import run_op_node
 
 
 def test_onehot():
-    runtime = get_runtime()
     param = ov.parameter([3], dtype=np.int32)
     model = ov.one_hot(param, 3, 1, 0, 0)
-    computation = runtime.computation(model, param)
-
-    expected = np.eye(3)[np.array([1, 0, 2])]
-    input_data = np.array([1, 0, 2], dtype=np.int32)
-    result = computation(input_data)
-    assert np.allclose(result, expected)
+    assert model.get_output_size() == 1
+    assert model.get_type_name() == "OneHot"
+    assert list(model.get_output_shape(0)) == [3, 3]
 
 
 def test_one_hot():
@@ -27,10 +21,11 @@ def test_one_hot():
     on_value = 5
     off_value = 10
     axis = -1
-    excepted = [[5, 10], [10, 5], [10, 10]]
 
-    result = run_op_node([data, depth, on_value, off_value], ov.one_hot, axis)
-    assert np.allclose(result, excepted)
+    node = ov.one_hot(data, depth, on_value, off_value, axis)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "OneHot"
+    assert list(node.get_output_shape(0)) == [3, 2]
 
 
 def test_range():
@@ -38,5 +33,7 @@ def test_range():
     stop = 35
     step = 5
 
-    result = run_op_node([start, stop, step], ov.range)
-    assert np.allclose(result, [5, 10, 15, 20, 25, 30])
+    node = ov.range(start, stop, step)
+    assert node.get_output_size() == 1
+    assert node.get_type_name() == "Range"
+    assert list(node.get_output_shape(0)) == [6]
diff --git a/src/bindings/python/tests/test_graph/test_utils.py b/src/bindings/python/tests/test_graph/test_utils.py
index 4801ea14a52..06f25d89adb 100644
--- a/src/bindings/python/tests/test_graph/test_utils.py
+++ b/src/bindings/python/tests/test_graph/test_utils.py
@@ -2,7 +2,6 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
-import numpy as np
 import openvino.runtime as ov
 import pytest
 from openvino._pyopenvino.util import deprecation_warning
diff --git a/src/bindings/python/tests/test_graph/util.py b/src/bindings/python/tests/test_graph/util.py
index a5aed287f5e..1487513ffae 100644
--- a/src/bindings/python/tests/test_graph/util.py
+++ b/src/bindings/python/tests/test_graph/util.py
@@ -2,75 +2,6 @@
 # Copyright (C) 2018-2022 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 
-from typing import Any, Callable, List, Union
-
-import numpy as np
-
-import openvino.runtime.opset8 as ov
-from openvino.runtime.utils.types import NumericData
-from tests.runtime import get_runtime
-from string import ascii_uppercase
-
-
-def _get_numpy_dtype(scalar):
-    return np.array([scalar]).dtype
-
-
-def run_op_node(input_data, op_fun, *args):
-    # type: (Union[NumericData, List[NumericData]], Callable, *Any) -> List[NumericData]
-    """Run computation on node performing `op_fun`.
-
-    `op_fun` has to accept a node as an argument.
-
-    This function converts passed raw input data to graph Constant Node and that form is passed
-    to `op_fun`.
-
-    :param input_data: The input data for performed computation.
-    :param op_fun: The function handler for operation we want to carry out.
-    :param args: The arguments passed to operation we want to carry out.
-    :return: The result from computations.
-    """
-    runtime = get_runtime()
-    comp_args = []
-    op_fun_args = []
-    comp_inputs = []
-
-    for idx, data in enumerate(input_data):
-        node = None
-        if np.isscalar(data):
-            node = ov.parameter([], name=ascii_uppercase[idx], dtype=_get_numpy_dtype(data))
-        else:
-            node = ov.parameter(data.shape, name=ascii_uppercase[idx], dtype=data.dtype)
-        op_fun_args.append(node)
-        comp_args.append(node)
-        comp_inputs.append(data)
-
-    op_fun_args.extend(args)
-    node = op_fun(*op_fun_args)
-    computation = runtime.computation(node, *comp_args)
-    return computation(*comp_inputs)
-
-
-def run_op_numeric_data(input_data, op_fun, *args):
-    # type: (NumericData, Callable, *Any) -> List[NumericData]
-    """Run computation on node performing `op_fun`.
-
-    `op_fun` has to accept a scalar or an array.
-
-    This function passess input data AS IS. This mean that in case they're a scalar (integral,
-    or floating point value) or a NumPy's ndarray object they will be automatically converted
-    to graph's Constant Nodes.
-
-    :param input_data: The input data for performed computation.
-    :param op_fun: The function handler for operation we want to carry out.
-    :param args: The arguments passed to operation we want to carry out.
-    :return: The result from computations.
-    """
-    runtime = get_runtime()
-    node = op_fun(input_data, *args)
-    computation = runtime.computation(node)
-    return computation()
-
 
 def count_ops_of_type(func, op_type):
     count = 0
diff --git a/src/bindings/python/tests_compatibility/test_ngraph/test_create_op.py b/src/bindings/python/tests_compatibility/test_ngraph/test_create_op.py
index 81d4b6e1bfa..7fec474850b 100644
--- a/src/bindings/python/tests_compatibility/test_ngraph/test_create_op.py
+++ b/src/bindings/python/tests_compatibility/test_ngraph/test_create_op.py
@@ -259,7 +259,7 @@ def test_deformable_psroi_pooling(dtype):
         ([2, 3, 5, 6], [7, 4], [7], 2, 2, 1, 1.0, "avg", "asymmetric", [7, 3, 2, 2]),
         ([10, 3, 5, 5], [7, 4], [7], 3, 4, 1, 1.0, "avg", "half_pixel_for_nn", [7, 3, 3, 4]),
         ([10, 3, 5, 5], [3, 4], [3], 3, 4, 1, 1.0, "avg", "half_pixel", [3, 3, 3, 4]),
-        ([10, 3, 5, 5], [3, 4], [3], 3, 4, 1, float(1), "avg", "half_pixel", [3, 3, 3, 4]),
+        ([10, 3, 5, 5], [3, 4], [3], 3, 4, 1, np.float32(1), "avg", "half_pixel", [3, 3, 3, 4]),
     ],
 )
 def test_roi_align(data_shape, rois, batch_indices, pooled_h, pooled_w, sampling_ratio, spatial_scale, mode, aligned_mode, expected_shape):
@@ -1882,11 +1882,11 @@ def test_multiclass_nms():
                            0.0, -0.1, 1.0, 0.9, 0.0, 10.0, 1.0, 11.0,
                            0.0, 10.1, 1.0, 11.1, 0.0, 100.0, 1.0, 101.0], dtype="float32")
     boxes_data = boxes_data.reshape([1, 6, 4])
-    box = ng.constant(boxes_data, dtype=float)
+    box = ng.constant(boxes_data, dtype=np.float32)
     scores_data = np.array([0.9, 0.75, 0.6, 0.95, 0.5, 0.3,
                             0.95, 0.75, 0.6, 0.80, 0.5, 0.3], dtype="float32")
     scores_data = scores_data.reshape([1, 2, 6])
-    score = ng.constant(scores_data, dtype=float)
+    score = ng.constant(scores_data, dtype=np.float32)
 
     nms_node = ng.multiclass_nms(box, score, None, output_type="i32", nms_top_k=3,
                                  iou_threshold=0.5, score_threshold=0.0, sort_result_type="classid",
@@ -1907,13 +1907,13 @@ def test_multiclass_nms():
                             [9.66, 3.36, 18.57, 13.26]],
                            [[6.50, 7.00, 13.33, 17.63],
                             [0.73, 5.34, 19.97, 19.97]]]).astype("float32")
-    box = ng.constant(boxes_data, dtype=float)
+    box = ng.constant(boxes_data, dtype=np.float32)
     scores_data = np.array([[0.34, 0.66],
                             [0.45, 0.61],
                             [0.39, 0.59]]).astype("float32")
-    score = ng.constant(scores_data, dtype=float)
+    score = ng.constant(scores_data, dtype=np.float32)
     rois_num_data = np.array([3]).astype("int32")
-    roisnum = ng.constant(rois_num_data, dtype=int)
+    roisnum = ng.constant(rois_num_data, dtype=np.int32)
     nms_node = ng.multiclass_nms(box, score, roisnum, output_type="i32", nms_top_k=3,
                                  iou_threshold=0.5, score_threshold=0.0, sort_result_type="classid",
                                  nms_eta=1.0)
@@ -1933,11 +1933,11 @@ def test_matrix_nms():
                            0.0, -0.1, 1.0, 0.9, 0.0, 10.0, 1.0, 11.0,
                            0.0, 10.1, 1.0, 11.1, 0.0, 100.0, 1.0, 101.0], dtype="float32")
     boxes_data = boxes_data.reshape([1, 6, 4])
-    box = ng.constant(boxes_data, dtype=float)
+    box = ng.constant(boxes_data, dtype=np.float32)
     scores_data = np.array([0.9, 0.75, 0.6, 0.95, 0.5, 0.3,
                             0.95, 0.75, 0.6, 0.80, 0.5, 0.3], dtype="float32")
     scores_data = scores_data.reshape([1, 2, 6])
-    score = ng.constant(scores_data, dtype=float)
+    score = ng.constant(scores_data, dtype=np.float32)
 
     nms_node = ng.matrix_nms(box, score, output_type="i32", nms_top_k=3,
                              score_threshold=0.0, sort_result_type="score", background_class=0,
@@ -2268,7 +2268,7 @@ def test_interpolate_opset10(dtype, expected_shape, shape_calculation_mode):
 
 def test_is_finite_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ng.parameter(input_shape, float, name="InputData")
+    input_node = ng.parameter(input_shape, np.float32, name="InputData")
     node = ng_opset10.is_finite(input_node)
 
     assert node.get_type_name() == "IsFinite"
@@ -2278,7 +2278,7 @@ def test_is_finite_opset10():
 
 def test_is_inf_opset10_default():
     input_shape = [2, 2, 2, 2]
-    input_node = ng.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ng.parameter(input_shape, dtype=np.float32, name="InputData")
     node = ng_opset10.is_inf(input_node)
 
     assert node.get_type_name() == "IsInf"
@@ -2292,7 +2292,7 @@ def test_is_inf_opset10_default():
 
 def test_is_inf_opset10_custom_attribute():
     input_shape = [2, 2, 2]
-    input_node = ng.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ng.parameter(input_shape, dtype=np.float32, name="InputData")
     attributes = {
         "detect_positive": False,
     }
@@ -2309,7 +2309,7 @@ def test_is_inf_opset10_custom_attribute():
 
 def test_is_inf_opset10_custom_all_attributes():
     input_shape = [2, 2, 2]
-    input_node = ng.parameter(input_shape, dtype=float, name="InputData")
+    input_node = ng.parameter(input_shape, dtype=np.float32, name="InputData")
     attributes = {
         "detect_negative": False,
         "detect_positive": True,
@@ -2327,7 +2327,7 @@ def test_is_inf_opset10_custom_all_attributes():
 
 def test_is_nan_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ng.parameter(input_shape, float, name="InputData")
+    input_node = ng.parameter(input_shape, np.float32, name="InputData")
     node = ng_opset10.is_nan(input_node)
 
     assert node.get_type_name() == "IsNaN"
@@ -2338,7 +2338,7 @@ def test_is_nan_opset10():
 
 def test_unique_opset10():
     input_shape = [1, 2, 3, 4]
-    input_node = ng.parameter(input_shape, float, name="input_data")
+    input_node = ng.parameter(input_shape, np.float32, name="input_data")
     axis = ng.constant([1], np.int32, [1])
 
     node = ng_opset10.unique(input_node, axis, False, "i32")
diff --git a/src/bindings/python/tests_compatibility/test_ngraph/test_if.py b/src/bindings/python/tests_compatibility/test_ngraph/test_if.py
index 7e69aeef8e3..7ed8a172aef 100644
--- a/src/bindings/python/tests_compatibility/test_ngraph/test_if.py
+++ b/src/bindings/python/tests_compatibility/test_ngraph/test_if.py
@@ -115,14 +115,14 @@ def simple_if_without_parameters(condition_val):
     condition = ng.constant(condition_val, dtype=bool)
 
     # then_body
-    then_constant = ng.constant(0.7, dtype=float)
+    then_constant = ng.constant(0.7, dtype=np.float32)
     then_body_res_1 = ng.result(then_constant)
     then_body = GraphBody([], [then_body_res_1])
     then_body_inputs = []
     then_body_outputs = [TensorIteratorBodyOutputDesc(0, 0)]
 
     # else_body
-    else_const = ng.constant(9.0, dtype=float)
+    else_const = ng.constant(9.0, dtype=np.float32)
     else_body_res_1 = ng.result(else_const)
     else_body = GraphBody([], [else_body_res_1])
     else_body_inputs = []
diff --git a/src/bindings/python/tests_compatibility/test_ngraph/test_ops_binary.py b/src/bindings/python/tests_compatibility/test_ngraph/test_ops_binary.py
index 3a410d01bf6..66c295e006b 100644
--- a/src/bindings/python/tests_compatibility/test_ngraph/test_ops_binary.py
+++ b/src/bindings/python/tests_compatibility/test_ngraph/test_ops_binary.py
@@ -80,7 +80,6 @@ def test_binary_logical_op(ng_api_helper):
     parameter_b = ng.parameter(shape, name="B", dtype=bool)
 
     model = ng_api_helper(parameter_a, parameter_b)
-
     assert model.get_output_size() == 1
     assert list(model.get_output_shape(0)) == [2, 2]
     assert model.get_output_element_type(0) == Type.boolean
diff --git a/src/bindings/python/tests_compatibility/test_ngraph/test_utils.py b/src/bindings/python/tests_compatibility/test_ngraph/test_utils.py
index 0178296e938..d8e3ded2d7f 100644
--- a/src/bindings/python/tests_compatibility/test_ngraph/test_utils.py
+++ b/src/bindings/python/tests_compatibility/test_ngraph/test_utils.py
@@ -7,7 +7,7 @@ from ngraph.impl import Shape
 
 
 def test_get_constant_from_source_success():
-    dtype = np.int
+    dtype = np.int32
     input1 = ng.parameter(Shape([5, 5]), dtype=dtype, name="input_1")
     input2 = ng.parameter(Shape([25]), dtype=dtype, name="input_2")
     shape_of = ng.shape_of(input2, name="shape_of")
@@ -19,7 +19,7 @@ def test_get_constant_from_source_success():
 
 
 def test_get_constant_from_source_failed():
-    dtype = np.int
+    dtype = np.int32
     input1 = ng.parameter(Shape([5, 5]), dtype=dtype, name="input_1")
     input2 = ng.parameter(Shape([1]), dtype=dtype, name="input_2")
     reshape = ng.reshape(input1, input2, special_zero=True)