diff --git a/unit_tests/recipies/Test_ConstantFolding.cpp b/unit_tests/recipies/Test_ConstantFolding.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..c4866b1258702b93a1bce80501d9acd094a65741
--- /dev/null
+++ b/unit_tests/recipies/Test_ConstantFolding.cpp
@@ -0,0 +1,85 @@
+/********************************************************************************
+ * Copyright (c) 2023 CEA-List
+ *
+ * This program and the accompanying materials are made available under the
+ * terms of the Eclipse Public License 2.0 which is available at
+ * http://www.eclipse.org/legal/epl-2.0.
+ *
+ * SPDX-License-Identifier: EPL-2.0
+ *
+ ********************************************************************************/
+
+#include <catch2/catch_test_macros.hpp>
+
+#include "aidge/recipes/Recipes.hpp"
+#include "aidge/operator/Add.hpp"
+#include "aidge/operator/MatMul.hpp"
+#include "aidge/operator/Producer.hpp"
+#include "aidge/graph/OpArgs.hpp"
+#include "aidge/scheduler/SequentialScheduler.hpp"
+#include "aidge/utils/TensorUtils.hpp"
+#include <cstddef>
+
+using namespace Aidge;
+
+TEST_CASE("[ConstantFolding] test") {
+    // generate the original GraphView
+    auto matmul0 = MatMul("matmul0");
+    auto add0 = Add(2, "add0");
+    auto matmul1 = MatMul("matmul1");
+    auto add1 = Add(2, "add1");
+
+    auto b0 = Producer(std::make_shared<Tensor>(Array1D<float,5>{{1, 2, 3, 4, 5}}), "B0", true);
+    auto w0 = Producer(std::make_shared<Tensor>(Array2D<float,5,5>{{{1, 2, 3, 4, 5}, {6, 7, 8, 9, 0}, {1, 2, 3, 4, 5}, {6, 7, 8, 9, 0}, {1, 2, 3, 4, 5}}}), "W0", true);
+    auto b1 = Producer(std::make_shared<Tensor>(Array1D<float,5>{{1, 2, 3, 4, 5}}), "B1", true);
+    auto w1 = Producer(std::make_shared<Tensor>(Array2D<float,5,5>{{{6, 7, 8, 9, 0}, {1, 2, 3, 4, 5}, {6, 7, 8, 9, 0}, {1, 2, 3, 4, 5}, {6, 7, 8, 9, 0}}}),"W1", true);
+    auto input = Producer(std::make_shared<Tensor>(Array2D<float,2,5>{{{1, 2, 3, 4, 5}, {6, 7, 8, 9, 0}}}), "input", true);
+
+    input->addChild(matmul0, 0, 0);
+    w0->addChild(matmul0, 0, 1);
+
+    matmul0->addChild(add0, 0, 0);
+    b0->addChild(add0, 0, 1);
+
+    add0->addChild(matmul1, 0, 0);
+    w1->addChild(matmul1, 0, 1);
+
+    matmul1->addChild(add1, 0, 0);
+    b1->addChild(add1, 0, 1);
+
+    auto g = std::make_shared<GraphView>();
+    g->add({input, w0, matmul0, b0, add0, w1, matmul1, b1, add1});
+    g->setBackend("cpu");
+    g->forwardDims();
+
+    // Check original graph
+    REQUIRE(g->getNodes() ==
+            std::set<std::shared_ptr<Node>>({input, w0, matmul0, b0, add0, w1, matmul1, b1, add1}));
+    REQUIRE(((matmul0->getParent(0) == input) && (matmul0->getParent(1) == w0)));
+    REQUIRE(((add0->getParent(0) == matmul0) && (add0->getParent(1) == b0)));
+    REQUIRE(((matmul1->getParent(0) == add0) && (matmul1->getParent(1) == w1)));
+    REQUIRE(((add1->getParent(0) == matmul1) && (add1->getParent(1) == b1)));
+
+    auto scheduler = SequentialScheduler(g);
+    scheduler.forward();
+
+    const std::shared_ptr<Tensor> result = std::make_shared<Tensor>(Array2D<float,2,5>{{
+        { 1201.000000, 1532.000000, 1863.000000, 2194.000000, 785.000000},
+        { 2501.000000, 3207.000000, 3913.000000, 4619.000000, 1735.000000}
+    }});
+
+    auto add1Op = std::static_pointer_cast<Add_Op>(add1->getOperator());
+    REQUIRE(approxEq<float>(*(add1Op->getOutput(0)), *result));
+
+	// Transform GraphView inplace
+    constantFolding(g);
+
+	// Check new GraphView
+	std::set<std::shared_ptr<Node>> newNodes = g->getNodes();
+	REQUIRE(newNodes != std::set<std::shared_ptr<Node>>({input, w0, matmul0, b0, add0, w1, matmul1, b1, add1}));
+	REQUIRE(newNodes.size() == 1);
+	REQUIRE((*newNodes.cbegin())->type() == "Producer");
+
+    auto prodOp = std::static_pointer_cast<Producer_Op>((*newNodes.cbegin())->getOperator());
+    REQUIRE(approxEq<float>(*(prodOp->getOutput(0)), *result));
+}