diff --git a/CMakeLists.txt b/CMakeLists.txt
index 55ecf8cd..81bf2215 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -40,12 +40,16 @@ add_definitions(${MLIR_DEFINITIONS})
 
 # Tools built by this project or LLVM
 set(MLIR_NEURA_OPT ${CMAKE_BINARY_DIR}/tools/mlir-neura-opt/mlir-neura-opt)
+set(NEURA_INTERPRETER ${CMAKE_BINARY_DIR}/tools/neura-interpreter/neura-interpreter)
 set(FILECHECK ${LLVM_TOOLS_BINARY_DIR}/FileCheck)
 set(MLIR_OPT ${LLVM_TOOLS_BINARY_DIR}/mlir-opt)
 set(MLIR_TRANSLATE ${LLVM_TOOLS_BINARY_DIR}/mlir-translate)
 set(LLC ${LLVM_TOOLS_BINARY_DIR}/llc)
 
-# Configure lit.cfg from lit.cfg.in
+# Builds the interpreter.
+add_subdirectory(tools/neura-interpreter)
+
+# Configures lit.cfg from lit.cfg.in
 configure_file(
   ${CMAKE_SOURCE_DIR}/test/lit.cfg.in
   #  ${CMAKE_BINARY_DIR}/test/lit.cfg
@@ -53,10 +57,11 @@ configure_file(
   @ONLY
 )
 
-# Add a custom target for running lit tests
+# Adds a custom target for running lit tests
 add_custom_target(check-neura
   COMMAND ${LLVM_EXTERNAL_LIT} -sv ${CMAKE_BINARY_DIR}/test
   DEPENDS mlir-neura-opt
   WORKING_DIRECTORY ${CMAKE_BINARY_DIR}
   COMMENT "Running Sora Dialect Tests with lit"
 )
+
diff --git a/test/lit.cfg.in b/test/lit.cfg.in
index c732a7d0..33640d31 100644
--- a/test/lit.cfg.in
+++ b/test/lit.cfg.in
@@ -9,6 +9,7 @@ config.test_exec_root = os.path.dirname(__file__)
 
 # Tool substitutions from CMake
 config.substitutions.append(('mlir-neura-opt', '@MLIR_NEURA_OPT@'))
+config.substitutions.append(('neura-interpreter', '@NEURA_INTERPRETER@'))
 config.substitutions.append(('FileCheck', '@FILECHECK@'))
 config.substitutions.append(('mlir-opt', '@MLIR_OPT@'))
 config.substitutions.append(('mlir-translate', '@MLIR_TRANSLATE@'))
diff --git a/test/neura/interpreter/interpreter.mlir b/test/neura/interpreter/interpreter.mlir
new file mode 100644
index 00000000..bce96544
--- /dev/null
+++ b/test/neura/interpreter/interpreter.mlir
@@ -0,0 +1,13 @@
+// RUN: neura-interpreter %s | FileCheck %s
+
+module {
+  func.func @test() -> f32 {
+    %arg0 = arith.constant 9.0 : f32
+    %cst = arith.constant 2.0 : f32
+    %0 = neura.mov %arg0 : f32 -> f32
+    %1 = neura.mov %cst : f32 -> f32
+    %2 = "neura.fadd"(%0, %1) : (f32, f32) -> f32
+    return %2 : f32
+    // CHECK: 1.1
+  }
+}
diff --git a/test/neura/interpreter/lower_and_interpreter.mlir b/test/neura/interpreter/lower_and_interpreter.mlir
new file mode 100644
index 00000000..c76db420
--- /dev/null
+++ b/test/neura/interpreter/lower_and_interpreter.mlir
@@ -0,0 +1,39 @@
+// RUN: mlir-opt %s \
+// RUN:   --convert-scf-to-cf \
+// RUN:   --convert-math-to-llvm \
+// RUN:   --convert-arith-to-llvm \
+// RUN:   --convert-func-to-llvm \
+// RUN:   --convert-cf-to-llvm \
+// RUN:   --reconcile-unrealized-casts \
+// RUN:   -o %t-lowered-to-llvm.mlir
+
+// RUN: mlir-translate -mlir-to-llvmir \
+// RUN:   %t-lowered-to-llvm.mlir \
+// RUN:   -o %t-lower_and_interpreter.ll
+
+// RUN: llc %t-lower_and_interpreter.ll \
+// RUN:   -filetype=obj -o %t-out.o
+
+// RUN: clang++ main.cpp %t-out.o \
+// RUN:   -o %t-out.bin
+
+// RUN: %t-out.bin > %t-dumped_output.txt
+
+// RUN: mlir-neura-opt --lower-arith-to-neura --insert-mov %s \
+// RUN:   -o %t-neura.mlir
+
+// RUN: neura-interpreter %t-neura.mlir >> %t-dumped_output.txt
+
+// RUN: FileCheck %s < %t-dumped_output.txt
+
+module {
+  func.func @test() -> f32 attributes { llvm.emit_c_interface }{
+    %arg0 = arith.constant 9.0 : f32
+    %cst = arith.constant 2.0 : f32
+    %0 = arith.addf %arg0, %cst : f32
+    // CHECK: Golden output: [[OUTPUT:[0-9]+\.[0-9]+]]
+    // CHECK: [neura-interpreter] Output: [[OUTPUT]]
+    return %0 : f32
+  }
+}
+
diff --git a/test/neura/interpreter/main.cpp b/test/neura/interpreter/main.cpp
new file mode 100644
index 00000000..0a67b549
--- /dev/null
+++ b/test/neura/interpreter/main.cpp
@@ -0,0 +1,10 @@
+#include <cstdio>
+
+extern "C" float test();
+
+int main() {
+  float result = test();
+  std::printf("Golden output: %f\n", result);
+  return 0;
+}
+
diff --git a/tools/neura-interpreter/CMakeLists.txt b/tools/neura-interpreter/CMakeLists.txt
new file mode 100644
index 00000000..ff9c62a2
--- /dev/null
+++ b/tools/neura-interpreter/CMakeLists.txt
@@ -0,0 +1,21 @@
+add_executable(neura-interpreter neura-interpreter.cpp)
+
+target_include_directories(neura-interpreter
+  PRIVATE
+    ${CMAKE_SOURCE_DIR}/include
+    ${CMAKE_BINARY_DIR}/include
+)
+
+target_link_libraries(neura-interpreter
+  PRIVATE
+    MLIRParser
+    MLIRIR
+    MLIRArithDialect
+    MLIRSupport
+    MLIRFuncDialect
+    MLIRDialect       # Optional: for builtin
+    NeuraDialect      # Your dialect
+)
+
+mlir_check_all_link_libraries(neura-interpreter)
+
diff --git a/tools/neura-interpreter/neura-interpreter.cpp b/tools/neura-interpreter/neura-interpreter.cpp
new file mode 100644
index 00000000..972e442c
--- /dev/null
+++ b/tools/neura-interpreter/neura-interpreter.cpp
@@ -0,0 +1,89 @@
+#include "llvm/Support/Format.h"
+#include "llvm/Support/SourceMgr.h"
+#include "mlir/Dialect/Arith/IR/Arith.h"
+#include "mlir/IR/MLIRContext.h"
+#include "mlir/IR/BuiltinOps.h"
+#include "mlir/IR/Operation.h"
+#include "mlir/Parser/Parser.h"
+#include "mlir/Support/FileUtilities.h"
+#include "mlir/Support/LogicalResult.h"
+#include "mlir/IR/AsmState.h"
+#include "mlir/InitAllDialects.h"
+#include "mlir/Dialect/Func/IR/FuncOps.h"
+
+#include "NeuraDialect/NeuraDialect.h"
+#include "NeuraDialect/NeuraOps.h"
+
+#include <unordered_map>
+#include <iostream>
+
+using namespace mlir;
+
+int main(int argc, char **argv) {
+  if (argc < 2) {
+    llvm::errs() << "Usage: neura-interpreter <input.mlir>\n";
+    return 1;
+  }
+
+  DialectRegistry registry;
+  registry.insert<neura::NeuraDialect, func::FuncDialect, arith::ArithDialect>();
+
+  MLIRContext context;
+  context.appendDialectRegistry(registry);
+
+  llvm::SourceMgr sourceMgr;
+  auto fileOrErr = mlir::openInputFile(argv[1]);
+  if (!fileOrErr) {
+    llvm::errs() << "Error opening file\n";
+    return 1;
+  }
+
+  sourceMgr.AddNewSourceBuffer(std::move(fileOrErr), llvm::SMLoc());
+
+  OwningOpRef<ModuleOp> module = parseSourceFile<ModuleOp>(sourceMgr, &context);
+  if (!module) {
+    llvm::errs() << "Failed to parse MLIR input file\n";
+    return 1;
+  }
+
+  llvm::DenseMap<Value, float> valueMap;
+
+  for (auto func : module->getOps<func::FuncOp>()) {
+    Block &block = func.getBody().front();
+
+    for (Operation &op : block.getOperations()) {
+      if (auto constOp = dyn_cast<mlir::arith::ConstantOp>(op)) {
+        auto attr = constOp.getValue();
+      
+        float val = 0.0f;
+      
+        if (auto floatAttr = llvm::dyn_cast<mlir::FloatAttr>(attr)) {
+          val = floatAttr.getValueAsDouble();  // or .convertToFloat()
+        } else if (auto intAttr = llvm::dyn_cast<mlir::IntegerAttr>(attr)) {
+          val = static_cast<float>(intAttr.getInt()); // interpret integer as float
+        } else {
+          llvm::errs() << "Unsupported constant type in arith.constant\n";
+          return 1;
+        }
+      
+        valueMap[constOp.getResult()] = val;
+      } else if (auto movOp = dyn_cast<neura::MovOp>(op)) {
+        valueMap[movOp.getResult()] = valueMap[movOp.getOperand()];
+      } else if (auto faddOp = dyn_cast<neura::FAddOp>(op)) {
+        float lhs = valueMap[faddOp.getLhs()];
+        float rhs = valueMap[faddOp.getRhs()];
+        valueMap[faddOp.getResult()] = lhs + rhs;
+      } else if (auto retOp = dyn_cast<func::ReturnOp>(op)) {
+        float result = valueMap[retOp.getOperand(0)];
+        llvm::outs() << "[neura-interpreter] Output: " << llvm::format("%.6f", result) << "\n";
+      } else {
+        llvm::errs() << "Unhandled op: ";
+        op.print(llvm::errs());
+        llvm::errs() << "\n";
+        return 1;
+      }
+    }
+  }
+
+  return 0;
+}