diff --git a/neuralnetworks/1.2/Android.bp b/neuralnetworks/1.2/Android.bp
index e155bbd..528a2c7 100644
--- a/neuralnetworks/1.2/Android.bp
+++ b/neuralnetworks/1.2/Android.bp
@@ -9,6 +9,9 @@
     srcs: [
         "types.hal",
         "IDevice.hal",
+        "IExecutionCallback.hal",
+        "IPreparedModel.hal",
+        "IPreparedModelCallback.hal",
     ],
     interfaces: [
         "android.hardware.neuralnetworks@1.0",
diff --git a/neuralnetworks/1.2/IDevice.hal b/neuralnetworks/1.2/IDevice.hal
index aff4cf3..6a77961 100644
--- a/neuralnetworks/1.2/IDevice.hal
+++ b/neuralnetworks/1.2/IDevice.hal
@@ -17,9 +17,9 @@
 package android.hardware.neuralnetworks@1.2;
 
 import @1.0::ErrorStatus;
-import @1.0::IPreparedModelCallback;
 import @1.1::ExecutionPreference;
 import @1.1::IDevice;
+import IPreparedModelCallback;
 
 /**
  * This interface represents a device driver.
diff --git a/neuralnetworks/1.2/IExecutionCallback.hal b/neuralnetworks/1.2/IExecutionCallback.hal
new file mode 100644
index 0000000..667e0d6
--- /dev/null
+++ b/neuralnetworks/1.2/IExecutionCallback.hal
@@ -0,0 +1,48 @@
+/*
+ * Copyright (C) 2018 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.2;
+
+import @1.0::ErrorStatus;
+import @1.0::IExecutionCallback;
+
+/**
+ * IExecutionCallback must be used to return the error status result from an
+ * execution asynchronously launched from IPreparedModel::execute.
+ */
+interface IExecutionCallback extends @1.0::IExecutionCallback {
+
+    /**
+     * Either notify_1_2 or notify must be invoked immediately after the asynchronous
+     * task has finished performing the execution. Either notify_1_2 or notify must be
+     * provided with the ErrorStatus from the execution. If the asynchronous task is
+     * not launched, either notify_1_2 or notify must be invoked with the appropriate
+     * error.
+     *
+     * @param status Error status returned from launching the asynchronous task
+     *               (if the launch fails) or from the asynchronous task itself
+     *               (if the launch succeeds). Must be:
+     *               - NONE if the asynchronous execution was successful
+     *               - DEVICE_UNAVAILABLE if driver is offline or busy
+     *               - GENERAL_FAILURE if the asynchronous task resulted in an
+     *                 unspecified error
+     *               - OUTPUT_INSUFFICIENT_SIZE if provided output buffer is
+     *                 not large enough to store the resultant values
+     *               - INVALID_ARGUMENT if one of the input arguments to
+     *                 prepareModel is invalid
+     */
+    oneway notify_1_2(ErrorStatus status);
+};
diff --git a/neuralnetworks/1.2/IPreparedModel.hal b/neuralnetworks/1.2/IPreparedModel.hal
new file mode 100644
index 0000000..5590487
--- /dev/null
+++ b/neuralnetworks/1.2/IPreparedModel.hal
@@ -0,0 +1,74 @@
+/*
+ * Copyright (C) 2018 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.2;
+
+import @1.0::ErrorStatus;
+import @1.0::IPreparedModel;
+import @1.0::Request;
+import IExecutionCallback;
+
+/**
+ * IPreparedModel describes a model that has been prepared for execution and
+ * is used to launch executions.
+ */
+interface IPreparedModel extends @1.0::IPreparedModel {
+    /**
+     * Launches an asynchronous execution on a prepared model.
+     *
+     * The execution is performed asynchronously with respect to the caller.
+     * execute_1_2 must verify the inputs to the function are correct. If there is
+     * an error, execute_1_2 must immediately invoke the callback with the
+     * appropriate ErrorStatus value, then return with the same ErrorStatus. If
+     * the inputs to the function are valid and there is no error, execute_1_2 must
+     * launch an asynchronous task to perform the execution in the background,
+     * and immediately return with ErrorStatus::NONE. If the asynchronous task
+     * fails to launch, execute_1_2 must immediately invoke the callback with
+     * ErrorStatus::GENERAL_FAILURE, then return with
+     * ErrorStatus::GENERAL_FAILURE.
+     *
+     * When the asynchronous task has finished its execution, it must
+     * immediately invoke the callback object provided as an input to the
+     * execute_1_2 function. This callback must be provided with the ErrorStatus of
+     * the execution.
+     *
+     * If the prepared model was prepared from a model wherein all
+     * tensor operands have fully specified dimensions, and the inputs
+     * to the function are valid, then the execution should launch
+     * and complete successfully (ErrorStatus::NONE). There must be
+     * no failure unless the device itself is in a bad state.
+     *
+     * Multiple threads can call the execute_1_2 function on the same IPreparedModel
+     * object concurrently with different requests.
+     *
+     * @param request The input and output information on which the prepared
+     *                model is to be executed.
+     * @param callback A callback object used to return the error status of
+     *                 the execution. The callback object's notify function must
+     *                 be called exactly once, even if the execution was
+     *                 unsuccessful.
+     * @return status Error status of the call, must be:
+     *                - NONE if task is successfully launched
+     *                - DEVICE_UNAVAILABLE if driver is offline or busy
+     *                - GENERAL_FAILURE if there is an unspecified error
+     *                - OUTPUT_INSUFFICIENT_SIZE if provided output buffer is
+     *                  not large enough to store the resultant values
+     *                - INVALID_ARGUMENT if one of the input arguments is
+     *                  invalid
+     */
+    execute_1_2(Request request, IExecutionCallback callback)
+        generates (ErrorStatus status);
+};
diff --git a/neuralnetworks/1.2/IPreparedModelCallback.hal b/neuralnetworks/1.2/IPreparedModelCallback.hal
new file mode 100644
index 0000000..d3830c6
--- /dev/null
+++ b/neuralnetworks/1.2/IPreparedModelCallback.hal
@@ -0,0 +1,55 @@
+/*
+ * Copyright (C) 2018 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.2;
+
+import @1.0::ErrorStatus;
+import @1.0::IPreparedModelCallback;
+import IPreparedModel;
+
+/**
+ * IPreparedModelCallback must be used to return a prepared model produced by an
+ * asynchronous task launched from IDevice::prepareModel.
+ */
+interface IPreparedModelCallback extends @1.0::IPreparedModelCallback {
+
+    /**
+     * Either notify_1_2 or notify must be invoked immediately after the asynchronous
+     * task holding this callback has finished preparing the model. If the model was
+     * successfully prepared, either notify_1_2 or notify must be invoked with
+     * ErrorStatus::NONE and the prepared model. If the model was not able to be
+     * successfully prepared, either notify_1_2 or notify must be invoked with the
+     * appropriate ErrorStatus and nullptr as the IPreparedModel. If the asynchronous
+     * task holding this callback fails to launch or if the model provided to
+     * IDevice::prepareModel is invalid, either notify_1_2 or notify must be invoked
+     * with the appropriate error as well as nullptr for the IPreparedModel.
+     *
+     * @param status Error status returned from the asynchronous model
+     *               preparation task; must be:
+     *               - NONE if the asynchronous task successfully prepared the
+     *                 model
+     *               - DEVICE_UNAVAILABLE if driver is offline or busy
+     *               - GENERAL_FAILURE if the asynchronous task resulted in an
+     *                 unspecified error
+     *               - INVALID_ARGUMENT if one of the input arguments to
+     *                 prepareModel is invalid
+     * @param preparedModel A model that has been asynchronously prepared for
+     *                      execution. If the model was unable to be prepared
+     *                      due to an error, nullptr must be passed in place of
+     *                      the IPreparedModel object.
+     */
+    oneway notify_1_2(ErrorStatus status, IPreparedModel preparedModel);
+};
diff --git a/neuralnetworks/1.2/vts/functional/GeneratedTests.cpp b/neuralnetworks/1.2/vts/functional/GeneratedTests.cpp
index 79d5a60..9bff09c 100644
--- a/neuralnetworks/1.2/vts/functional/GeneratedTests.cpp
+++ b/neuralnetworks/1.2/vts/functional/GeneratedTests.cpp
@@ -40,8 +40,8 @@
 namespace vts {
 namespace functional {
 
-using ::android::hardware::neuralnetworks::V1_0::implementation::ExecutionCallback;
-using ::android::hardware::neuralnetworks::V1_0::implementation::PreparedModelCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::ExecutionCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::PreparedModelCallback;
 using ::android::nn::allocateSharedMemory;
 using ::test_helper::MixedTypedExample;
 
diff --git a/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_0.cpp b/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_0.cpp
index 42e22b0..56a61d4 100644
--- a/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_0.cpp
+++ b/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_0.cpp
@@ -40,8 +40,8 @@
 namespace vts {
 namespace functional {
 
-using ::android::hardware::neuralnetworks::V1_0::implementation::ExecutionCallback;
-using ::android::hardware::neuralnetworks::V1_0::implementation::PreparedModelCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::ExecutionCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::PreparedModelCallback;
 using ::android::nn::allocateSharedMemory;
 using ::test_helper::MixedTypedExample;
 
diff --git a/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_1.cpp b/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_1.cpp
index aab5cb6..1c781ec 100644
--- a/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_1.cpp
+++ b/neuralnetworks/1.2/vts/functional/GeneratedTestsV1_1.cpp
@@ -40,8 +40,8 @@
 namespace vts {
 namespace functional {
 
-using ::android::hardware::neuralnetworks::V1_0::implementation::ExecutionCallback;
-using ::android::hardware::neuralnetworks::V1_0::implementation::PreparedModelCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::ExecutionCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::PreparedModelCallback;
 using ::android::nn::allocateSharedMemory;
 using ::test_helper::MixedTypedExample;
 
diff --git a/neuralnetworks/1.2/vts/functional/ValidateModel.cpp b/neuralnetworks/1.2/vts/functional/ValidateModel.cpp
index 562ac33..4e0f2ff 100644
--- a/neuralnetworks/1.2/vts/functional/ValidateModel.cpp
+++ b/neuralnetworks/1.2/vts/functional/ValidateModel.cpp
@@ -25,15 +25,14 @@
 namespace neuralnetworks {
 namespace V1_2 {
 
-using V1_0::IPreparedModel;
 using V1_0::OperandLifeTime;
 using V1_1::ExecutionPreference;
 
 namespace vts {
 namespace functional {
 
-using ::android::hardware::neuralnetworks::V1_0::implementation::ExecutionCallback;
-using ::android::hardware::neuralnetworks::V1_0::implementation::PreparedModelCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::ExecutionCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::PreparedModelCallback;
 
 ///////////////////////// UTILITY FUNCTIONS /////////////////////////
 
@@ -62,7 +61,7 @@
     preparedModelCallback->wait();
     ErrorStatus prepareReturnStatus = preparedModelCallback->getStatus();
     ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, prepareReturnStatus);
-    sp<IPreparedModel> preparedModel = preparedModelCallback->getPreparedModel();
+    sp<IPreparedModel> preparedModel = getPreparedModel_1_2(preparedModelCallback);
     ASSERT_EQ(nullptr, preparedModel.get());
 }
 
diff --git a/neuralnetworks/1.2/vts/functional/ValidateRequest.cpp b/neuralnetworks/1.2/vts/functional/ValidateRequest.cpp
index b663535..e2722aa 100644
--- a/neuralnetworks/1.2/vts/functional/ValidateRequest.cpp
+++ b/neuralnetworks/1.2/vts/functional/ValidateRequest.cpp
@@ -33,8 +33,8 @@
 namespace vts {
 namespace functional {
 
-using ::android::hardware::neuralnetworks::V1_0::implementation::ExecutionCallback;
-using ::android::hardware::neuralnetworks::V1_0::implementation::PreparedModelCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::ExecutionCallback;
+using ::android::hardware::neuralnetworks::V1_2::implementation::PreparedModelCallback;
 using ::android::hidl::memory::V1_0::IMemory;
 using test_helper::for_all;
 using test_helper::MixedTyped;
@@ -68,7 +68,7 @@
     // retrieve prepared model
     preparedModelCallback->wait();
     ErrorStatus prepareReturnStatus = preparedModelCallback->getStatus();
-    *preparedModel = preparedModelCallback->getPreparedModel();
+    *preparedModel = getPreparedModel_1_2(preparedModelCallback);
 
     // The getSupportedOperations_1_2 call returns a list of operations that are
     // guaranteed not to fail if prepareModel_1_2 is called, and
@@ -101,7 +101,8 @@
 
     sp<ExecutionCallback> executionCallback = new ExecutionCallback();
     ASSERT_NE(nullptr, executionCallback.get());
-    Return<ErrorStatus> executeLaunchStatus = preparedModel->execute(request, executionCallback);
+    Return<ErrorStatus> executeLaunchStatus =
+        preparedModel->execute_1_2(request, executionCallback);
     ASSERT_TRUE(executeLaunchStatus.isOk());
     ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, static_cast<ErrorStatus>(executeLaunchStatus));
 
diff --git a/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.cpp b/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.cpp
index 90a910c..4eced82 100644
--- a/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.cpp
+++ b/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.cpp
@@ -58,6 +58,12 @@
     ::testing::VtsHalHidlTargetTestBase::TearDown();
 }
 
+sp<IPreparedModel> getPreparedModel_1_2(
+    const sp<V1_2::implementation::PreparedModelCallback>& callback) {
+    sp<V1_0::IPreparedModel> preparedModelV1_0 = callback->getPreparedModel();
+    return V1_2::IPreparedModel::castFrom(preparedModelV1_0).withDefault(nullptr);
+}
+
 }  // namespace functional
 }  // namespace vts
 
diff --git a/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.h b/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.h
index a87d788..dedab8d 100644
--- a/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.h
+++ b/neuralnetworks/1.2/vts/functional/VtsHalNeuralnetworks.h
@@ -17,6 +17,8 @@
 #ifndef VTS_HAL_NEURALNETWORKS_V1_2_H
 #define VTS_HAL_NEURALNETWORKS_V1_2_H
 
+#include "Callbacks.h"
+
 #include <android/hardware/neuralnetworks/1.0/types.h>
 #include <android/hardware/neuralnetworks/1.1/types.h>
 #include <android/hardware/neuralnetworks/1.2/IDevice.h>
@@ -77,6 +79,10 @@
 // Tag for the generated tests
 class GeneratedTest : public NeuralnetworksHidlTest {};
 
+// Utility function to get PreparedModel from callback and downcast to V1_2.
+sp<IPreparedModel> getPreparedModel_1_2(
+    const sp<V1_2::implementation::PreparedModelCallback>& callback);
+
 }  // namespace functional
 }  // namespace vts