diff --git a/current.txt b/current.txt
index 8d14fa3..8dfe34e 100644
--- a/current.txt
+++ b/current.txt
@@ -673,11 +673,11 @@
 df9c79c4fdde2821550c6d5c3d07f5ec0adfb1b702561ce543c906ddef698703 android.hardware.media.c2@1.1::IComponent
 a3eddd9bbdc87e8c22764070037dd1154f1cf006e6fba93364c4f85d4c134a19 android.hardware.media.c2@1.1::IComponentStore
 65c16331e57f6dd68b3971f06f78fe9e3209afb60630c31705aa355f9a52bf0d android.hardware.neuralnetworks@1.3::IBuffer
-9db064ee44268a876be0367ff771e618362d39ec603b6ecab17e1575725fcd87 android.hardware.neuralnetworks@1.3::IDevice
-4167dc3ad35e9cd0d2057d4868c7675ae2c3c9d05bbd614c1f5dccfa5fd68797 android.hardware.neuralnetworks@1.3::IExecutionCallback
-2fa3679ad7c94b5e88724adcd560c561041068a4ca565c63830e68101988746a android.hardware.neuralnetworks@1.3::IFencedExecutionCallback
-43088ffc71945b463a7279262cfe2e290f6ed2f15d3fd6032798a3be299fb08f android.hardware.neuralnetworks@1.3::IPreparedModel
-0439a1fbbec7f16e5e4c653d85ac685d51bfafbae15b8f8cca530acdd7d6a8ce android.hardware.neuralnetworks@1.3::IPreparedModelCallback
+278817920bfd5292a7713f97f1832cca53de3de640f7670e413d97c6e7fd581c android.hardware.neuralnetworks@1.3::IDevice
+127ba11efb8220dc3aec9a8f441b59eaf1c68d7f03f577833e1824de75a36b17 android.hardware.neuralnetworks@1.3::IExecutionCallback
+6e904be0ddca5ae1de8eba020e6c38ed935ea7d80cd08f47787f137a0ca58555 android.hardware.neuralnetworks@1.3::IFencedExecutionCallback
+2b0b10d2ea7a18a4048cd0eb83d35c19a817aeee95f65807fc31f4ef21381397 android.hardware.neuralnetworks@1.3::IPreparedModel
+eee3430cc86c97c7b407495863d8fb61da6f1a64b7721e77b9b4909b11b174e9 android.hardware.neuralnetworks@1.3::IPreparedModelCallback
 dd39887aa4fb60ce60ea9cc043edeadbbae6e922d09d3946311b0b410024ae14 android.hardware.neuralnetworks@1.3::types
 3e01d4446cd69fd1c48f8572efd97487bc179564b32bd795800b97bbe10be37b android.hardware.wifi@1.4::IWifi
 c67aaf26a7a40d14ea61e70e20afacbd0bb906df1704d585ac8599fbb69dd44b android.hardware.wifi.hostapd@1.2::IHostapd
diff --git a/neuralnetworks/1.3/IDevice.hal b/neuralnetworks/1.3/IDevice.hal
index 79f9c32..e0b04a8 100644
--- a/neuralnetworks/1.3/IDevice.hal
+++ b/neuralnetworks/1.3/IDevice.hal
@@ -48,19 +48,6 @@
     getCapabilities_1_3() generates (ErrorStatus status, Capabilities capabilities);
 
     /**
-     * Returns whether the device is able to complete or abort a task within a
-     * specified duration.
-     *
-     * @return prepareModelDeadline 'true' if the device supports completing or
-     *     aborting model preparation by the deadline when the deadline is supplied,
-     *     'false' otherwise.
-     * @return executionDeadline 'true' if the device supports completing or
-     *     aborting an execution by the deadline when the deadline is supplied,
-     *     'false' otherwise.
-     */
-    supportsDeadlines() generates (bool prepareModelDeadline, bool executionDeadline);
-
-    /**
      * Gets the supported operations in a model.
      *
      * getSupportedOperations indicates which operations of the top-level
@@ -140,14 +127,10 @@
      *
      * prepareModel_1_3 can be called with an optional deadline. If the model
      * is not able to be prepared before the provided deadline, the model
-     * preparation must be aborted, and either {@link
+     * preparation may be aborted, and either {@link
      * ErrorStatus::MISSED_DEADLINE_TRANSIENT} or {@link
-     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} must be returned. The error due
+     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The error due
      * to an abort must be sent the same way as other errors, described above.
-     * If the service reports that it does not support preparation deadlines via
-     * IDevice::supportsDeadlines, and prepareModel_1_3 is called with a
-     * deadline, then the argument is invalid, and {@link
-     * ErrorStatus::INVALID_ARGUMENT} must be returned.
      *
      * Optionally, the driver may save the prepared model to cache during the
      * asynchronous preparation. Any error that occurs when saving to cache must
@@ -172,9 +155,9 @@
      *     model.
      * @param priority The priority of the prepared model relative to other
      *     prepared models owned by the client.
-     * @param deadline The time by which the model must be prepared. If the
-     *     model cannot be prepared by the deadline, the preparation must be
-     *     aborted.
+     * @param deadline The time by which the model is expected to be prepared.
+     *     If the model cannot be prepared by the deadline, the preparation may
+     *     be aborted.
      * @param modelCache A vector of handles with each entry holding exactly one
      *     cache file descriptor for the security-sensitive cache. The length of
      *     the vector must either be 0 indicating that caching information is
@@ -209,8 +192,8 @@
      *     - GENERAL_FAILURE if there is an unspecified error
      *     - INVALID_ARGUMENT if one of the input arguments related to preparing
      *       the model is invalid
-     *     - MISSED_DEADLINE_* if the deadline for preparing a model cannot be
-     *       met
+     *     - MISSED_DEADLINE_* if the preparation is aborted because the model
+     *       cannot be prepared by the deadline
      *     - RESOURCE_EXHAUSTED_* if the task was aborted by the driver
      */
     prepareModel_1_3(Model model, ExecutionPreference preference,
@@ -262,14 +245,11 @@
      *
      * prepareModelFromCache_1_3 can be called with an optional deadline. If the
      * model is not able to prepared before the provided deadline, the model
-     * preparation must be aborted, and either {@link
+     * preparation may be aborted, and either {@link
      * ErrorStatus::MISSED_DEADLINE_TRANSIENT}
-     * or {@link ErrorStatus::MISSED_DEADLINE_PERSISTENT} must be returned. The
+     * or {@link ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The
      * error due to an abort must be sent the same way as other errors,
-     * described above. If the service reports that it does not support
-     * preparation deadlines via IDevice::supportsDeadlines, and
-     * prepareModelFromCache_1_3 is called with a deadline, then the argument is
-     * invalid, and {@link ErrorStatus::INVALID_ARGUMENT} must be returned.
+     * described above.
      *
      * The only information that may be unknown to the model at this stage is
      * the shape of the tensors, which may only be known at execution time. As
@@ -279,9 +259,9 @@
      * used with different shapes of inputs on different (possibly concurrent)
      * executions.
      *
-     * @param deadline The time by which the model must be prepared. If the
-     *     model cannot be prepared by the deadline, the preparation must be
-     *     aborted.
+     * @param deadline The time by which the model is expected to be prepared.
+     *     If the model cannot be prepared by the deadline, the preparation may
+     *     be aborted.
      * @param modelCache A vector of handles with each entry holding exactly one
      *     cache file descriptor for the security-sensitive cache. The length of
      *     the vector must match the numModelCache returned from getNumberOfCacheFilesNeeded.
@@ -307,8 +287,8 @@
      *     - GENERAL_FAILURE if caching is not supported or if there is an
      *       unspecified error
      *     - INVALID_ARGUMENT if one of the input arguments is invalid
-     *     - MISSED_DEADLINE_* if the deadline for preparing a model cannot be
-     *       met
+     *     - MISSED_DEADLINE_* if the preparation is aborted because the model
+     *       cannot be prepared by the deadline
      *     - RESOURCE_EXHAUSTED_* if the task was aborted by the driver
      */
     prepareModelFromCache_1_3(OptionalTimePoint deadline,
diff --git a/neuralnetworks/1.3/IExecutionCallback.hal b/neuralnetworks/1.3/IExecutionCallback.hal
index 439428a..ea11b17 100644
--- a/neuralnetworks/1.3/IExecutionCallback.hal
+++ b/neuralnetworks/1.3/IExecutionCallback.hal
@@ -47,7 +47,8 @@
      *                 corresponding output
      *               - INVALID_ARGUMENT if one of the input arguments to
      *                 prepareModel is invalid
-     *               - MISSED_DEADLINE_* if the deadline could not be met
+     *               - MISSED_DEADLINE_* if the execution is aborted because it
+     *                 cannot be completed by the deadline
      *               - RESOURCE_EXHAUSTED_* if the task was aborted by the driver
      * @param outputShapes A list of shape information of model output operands.
      *                     The index into "outputShapes" corresponds with to index
diff --git a/neuralnetworks/1.3/IFencedExecutionCallback.hal b/neuralnetworks/1.3/IFencedExecutionCallback.hal
index 6030809..949438e 100644
--- a/neuralnetworks/1.3/IFencedExecutionCallback.hal
+++ b/neuralnetworks/1.3/IFencedExecutionCallback.hal
@@ -38,8 +38,8 @@
      *                - DEVICE_UNAVAILABLE if driver is offline or busy
      *                - GENERAL_FAILURE if the asynchronous task resulted in an
      *                  unspecified error
-     *                - MISSED_DEADLINE_* if the deadline for executing a model
-     *                  cannot be met
+     *                - MISSED_DEADLINE_* if the execution is aborted because it
+     *                  cannot be completed by the deadline
      *                - RESOURCE_EXHAUSTED_* if the task was aborted by the
      *                  driver
      * @return timingLaunched The duration starts when executeFenced is called and ends when
diff --git a/neuralnetworks/1.3/IPreparedModel.hal b/neuralnetworks/1.3/IPreparedModel.hal
index 4ce3691..a1814b5 100644
--- a/neuralnetworks/1.3/IPreparedModel.hal
+++ b/neuralnetworks/1.3/IPreparedModel.hal
@@ -70,14 +70,10 @@
      *
      * execute_1_3 can be called with an optional deadline. If the execution
      * is not able to be completed before the provided deadline, the execution
-     * must be aborted, and either {@link
+     * may be aborted, and either {@link
      * ErrorStatus::MISSED_DEADLINE_TRANSIENT} or {@link
-     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} must be returned. The error due
+     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The error due
      * to an abort must be sent the same way as other errors, described above.
-     * If the service reports that it does not support execution deadlines via
-     * IDevice::supportsDeadlines, and execute_1_3 is called with a deadline,
-     * then the argument is invalid, and {@link ErrorStatus::INVALID_ARGUMENT}
-     * must be returned.
      *
      * Any number of calls to the execute* and executeSynchronously* functions,
      * in any combination, may be made concurrently, even on the same
@@ -89,9 +85,9 @@
      *                The duration runs from the time the driver sees the call
      *                to the execute_1_3 function to the time the driver invokes
      *                the callback.
-     * @param deadline The time by which the execution must complete. If the
-     *                 execution cannot be finished by the deadline, the
-     *                 execution must be aborted.
+     * @param deadline The time by which the execution is expected to complete.
+     *                 If the execution cannot be completed by the deadline, the
+     *                 execution may be aborted.
      * @param loopTimeoutDuration The maximum amount of time that should be spent
      *                            executing a {@link OperationType::WHILE}
      *                            operation. If a loop condition model does not
@@ -116,8 +112,8 @@
      *                  not large enough to store the resultant values
      *                - INVALID_ARGUMENT if one of the input arguments is
      *                  invalid
-     *                - MISSED_DEADLINE_* if the deadline for executing a model
-     *                  cannot be met
+     *                - MISSED_DEADLINE_* if the execution is aborted because it
+     *                  cannot be completed by the deadline
      *                - RESOURCE_EXHAUSTED_* if the task was aborted by the
      *                  driver
      */
@@ -150,16 +146,12 @@
      * (ErrorStatus::NONE): There must be no failure unless the device itself is
      * in a bad state.
      *
-     * executeSynchronously_1_3 can be called with an optional deadline. If the
+     * executeSynchronously_1_3 may be called with an optional deadline. If the
      * execution is not able to be completed before the provided deadline, the
-     * execution must be aborted, and either {@link
+     * execution may be aborted, and either {@link
      * ErrorStatus::MISSED_DEADLINE_TRANSIENT} or {@link
-     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} must be returned. The error due
+     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The error due
      * to an abort must be sent the same way as other errors, described above.
-     * If the service reports that it does not support execution deadlines via
-     * IDevice::supportsDeadlines, and executeSynchronously_1_3 is called with a
-     * deadline, then the argument is invalid, and
-     * {@link ErrorStatus::INVALID_ARGUMENT} must be returned.
      *
      * Any number of calls to the execute* and executeSynchronously* functions,
      * in any combination, may be made concurrently, even on the same
@@ -171,9 +163,9 @@
      *                The duration runs from the time the driver sees the call
      *                to the executeSynchronously_1_3 function to the time the driver
      *                returns from the function.
-     * @param deadline The time by which the execution must complete. If the
-     *                 execution cannot be finished by the deadline, the
-     *                 execution must be aborted.
+     * @param deadline The time by which the execution is expected to complete.
+     *                 If the execution cannot be finished by the deadline, the
+     *                 execution may be aborted.
      * @param loopTimeoutDuration The maximum amount of time that should be spent
      *                            executing a {@link OperationType::WHILE}
      *                            operation. If a loop condition model does not
@@ -194,8 +186,8 @@
      *                  corresponding output
      *                - INVALID_ARGUMENT if one of the input arguments is
      *                  invalid
-     *                - MISSED_DEADLINE_* if the deadline for executing a model
-     *                  cannot be met
+     *                - MISSED_DEADLINE_* if the execution is aborted because it
+     *                  cannot be completed by the deadline
      *                - RESOURCE_EXHAUSTED_* if the task was aborted by the
      *                  driver
      * @return outputShapes A list of shape information of model output operands.
@@ -236,17 +228,13 @@
      * any data object referenced by 'request' (described by the
      * {@link @1.0::DataLocation} of a {@link @1.0::RequestArgument}).
      *
-     * executeFenced can be called with an optional deadline and an optional duration.
+     * executeFenced may be called with an optional deadline and an optional duration.
      * If the execution is not able to be completed before the provided deadline or
      * within the timeout duration (measured from when all sync fences in waitFor are
-     * signaled), whichever comes earlier, the execution must be aborted, and either
+     * signaled), whichever comes earlier, the execution may be aborted, and either
      * {@link ErrorStatus::MISSED_DEADLINE_TRANSIENT} or {@link
-     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} must be returned. The error due
+     * ErrorStatus::MISSED_DEADLINE_PERSISTENT} may be returned. The error due
      * to an abort must be sent the same way as other errors, described above.
-     * If the service reports that it does not support execution deadlines via
-     * IDevice::supportsDeadlines, and executeFenced is called with a
-     * deadline or duration, then the argument is invalid, and
-     * {@link ErrorStatus::INVALID_ARGUMENT} must be returned.
      *
      * If any of the sync fences in waitFor changes to error status after the executeFenced
      * call succeeds, or the execution is aborted because it cannot finish before the deadline
@@ -263,9 +251,9 @@
      * @param waitFor A vector of sync fence file descriptors.
      *                Execution must not start until all sync fences have been signaled.
      * @param measure Specifies whether or not to measure duration of the execution.
-     * @param deadline The time by which the execution must complete. If the
-     *                 execution cannot be finished by the deadline, the
-     *                 execution must be aborted.
+     * @param deadline The time by which the execution is expected to complete.
+     *                 If the execution cannot be finished by the deadline, the
+     *                 execution may be aborted.
      * @param loopTimeoutDuration The maximum amount of time that should be spent
      *                            executing a {@link OperationType::WHILE}
      *                            operation. If a loop condition model does not
@@ -277,18 +265,18 @@
      *                            LoopTimeoutDurationNs::DEFAULT}. When
      *                            provided, the duration must not exceed {@link
      *                            LoopTimeoutDurationNs::MAXIMUM}.
-     * @param duration The length of time within which the execution must
-     *                 complete after all sync fences in waitFor are signaled. If the
-     *                 execution cannot be finished within the duration, the execution
-     *                 must be aborted.
+     * @param duration The length of time within which the execution is expected
+     *                 to complete after all sync fences in waitFor are signaled.
+     *                 If the execution cannot be finished within the duration,
+     *                 the execution may be aborted.
      * @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
      *                - INVALID_ARGUMENT if one of the input arguments is invalid, including
      *                                   fences in error states.
-     *                - MISSED_DEADLINE_* if the deadline for executing a model
-     *                  cannot be met
+     *                - MISSED_DEADLINE_* if the execution is aborted because it
+     *                  cannot be completed by the deadline
      *                - RESOURCE_EXHAUSTED_* if the task was aborted by the
      *                  driver
      * @return syncFence The sync fence that will be signaled when the task is completed.
diff --git a/neuralnetworks/1.3/IPreparedModelCallback.hal b/neuralnetworks/1.3/IPreparedModelCallback.hal
index 11ebbf4..c0d3416 100644
--- a/neuralnetworks/1.3/IPreparedModelCallback.hal
+++ b/neuralnetworks/1.3/IPreparedModelCallback.hal
@@ -47,8 +47,8 @@
      *                 unspecified error
      *               - INVALID_ARGUMENT if one of the input arguments to
      *                 prepareModel is invalid
-     *               - MISSED_DEADLINE_* if the deadline for executing a model
-     *                 cannot be met
+     *               - MISSED_DEADLINE_* if the preparation is aborted because
+     *                 the model cannot be prepared by the deadline
      *               - RESOURCE_EXHAUSTED_* if the task was aborted by the
      *                 driver
      * @param preparedModel A model that has been asynchronously prepared for
diff --git a/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.cpp b/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.cpp
index b04abe2..8c9393b 100644
--- a/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.cpp
+++ b/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.cpp
@@ -858,12 +858,6 @@
 void GeneratedTestBase::SetUp() {
     testing::TestWithParam<GeneratedTestParam>::SetUp();
     ASSERT_NE(kDevice, nullptr);
-
-    const Return<void> ret =
-            kDevice->supportsDeadlines([this](bool prepareModelDeadline, bool executionDeadline) {
-                mSupportsDeadlines = {prepareModelDeadline, executionDeadline};
-            });
-    ASSERT_TRUE(ret.isOk());
 }
 
 std::vector<NamedModel> getNamedModels(const FilterFn& filter) {
diff --git a/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.h b/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.h
index a8db515..834d335 100644
--- a/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.h
+++ b/neuralnetworks/1.3/vts/functional/GeneratedTestHarness.h
@@ -36,7 +36,6 @@
     void SetUp() override;
     const sp<IDevice> kDevice = getData(std::get<NamedDevice>(GetParam()));
     const test_helper::TestModel& kTestModel = *getData(std::get<NamedModel>(GetParam()));
-    std::pair<bool, bool> mSupportsDeadlines;
 };
 
 using FilterFn = std::function<bool(const test_helper::TestModel&)>;
diff --git a/neuralnetworks/1.3/vts/functional/QualityOfServiceTests.cpp b/neuralnetworks/1.3/vts/functional/QualityOfServiceTests.cpp
index fccc612..2663500 100644
--- a/neuralnetworks/1.3/vts/functional/QualityOfServiceTests.cpp
+++ b/neuralnetworks/1.3/vts/functional/QualityOfServiceTests.cpp
@@ -34,45 +34,52 @@
 using HidlToken =
         hidl_array<uint8_t, static_cast<uint32_t>(V1_2::Constant::BYTE_SIZE_OF_CACHE_TOKEN)>;
 
-enum class DeadlineBoundType { NOW, UNLIMITED };
-constexpr std::array<DeadlineBoundType, 2> deadlineBounds = {DeadlineBoundType::NOW,
-                                                             DeadlineBoundType::UNLIMITED};
+enum class DeadlineBoundType { NOW, UNLIMITED, SHORT };
+constexpr std::array<DeadlineBoundType, 3> deadlineBounds = {
+        DeadlineBoundType::NOW, DeadlineBoundType::UNLIMITED, DeadlineBoundType::SHORT};
 std::string toString(DeadlineBoundType type) {
     switch (type) {
         case DeadlineBoundType::NOW:
             return "NOW";
         case DeadlineBoundType::UNLIMITED:
             return "UNLIMITED";
+        case DeadlineBoundType::SHORT:
+            return "SHORT";
     }
     LOG(FATAL) << "Unrecognized DeadlineBoundType: " << static_cast<int>(type);
     return {};
 }
 
+constexpr auto kShortDuration = std::chrono::milliseconds{5};
+
 using Results = std::tuple<ErrorStatus, hidl_vec<OutputShape>, Timing>;
 using MaybeResults = std::optional<Results>;
 
 using ExecutionFunction =
         std::function<MaybeResults(const sp<IPreparedModel>& preparedModel, const Request& request,
-                                   DeadlineBoundType deadlineBound)>;
+                                   const OptionalTimePoint& deadline)>;
 
-static OptionalTimePoint makeOptionalTimePoint(DeadlineBoundType deadlineBoundType) {
-    OptionalTimePoint deadline;
+static OptionalTimePoint makeDeadline(DeadlineBoundType deadlineBoundType) {
+    const auto getNanosecondsSinceEpoch = [](const auto& time) -> uint64_t {
+        const auto timeSinceEpoch = time.time_since_epoch();
+        return std::chrono::duration_cast<std::chrono::nanoseconds>(timeSinceEpoch).count();
+    };
+
+    std::chrono::steady_clock::time_point timePoint;
     switch (deadlineBoundType) {
-        case DeadlineBoundType::NOW: {
-            const auto currentTime = std::chrono::steady_clock::now();
-            const auto currentTimeInNanoseconds =
-                    std::chrono::time_point_cast<std::chrono::nanoseconds>(currentTime);
-            const uint64_t nanosecondsSinceEpoch =
-                    currentTimeInNanoseconds.time_since_epoch().count();
-            deadline.nanosecondsSinceEpoch(nanosecondsSinceEpoch);
-        } break;
-        case DeadlineBoundType::UNLIMITED: {
-            const auto maxTime = std::chrono::time_point<std::chrono::steady_clock,
-                                                         std::chrono::nanoseconds>::max();
-            const uint64_t nanosecondsSinceEpoch = maxTime.time_since_epoch().count();
-            deadline.nanosecondsSinceEpoch(nanosecondsSinceEpoch);
-        } break;
+        case DeadlineBoundType::NOW:
+            timePoint = std::chrono::steady_clock::now();
+            break;
+        case DeadlineBoundType::UNLIMITED:
+            timePoint = std::chrono::steady_clock::time_point::max();
+            break;
+        case DeadlineBoundType::SHORT:
+            timePoint = std::chrono::steady_clock::now() + kShortDuration;
+            break;
     }
+
+    OptionalTimePoint deadline;
+    deadline.nanosecondsSinceEpoch(getNanosecondsSinceEpoch(timePoint));
     return deadline;
 }
 
@@ -80,7 +87,7 @@
                          std::optional<DeadlineBoundType> deadlineBound) {
     OptionalTimePoint deadline;
     if (deadlineBound.has_value()) {
-        deadline = makeOptionalTimePoint(deadlineBound.value());
+        deadline = makeDeadline(deadlineBound.value());
     }
 
     // see if service can handle model
@@ -131,7 +138,8 @@
                 // deadline has already passed when the driver would launch the
                 // execution. In this case, the driver must return
                 // MISSED_DEADLINE_*.
-                EXPECT_TRUE(prepareReturnStatus == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
+                EXPECT_TRUE(prepareReturnStatus == ErrorStatus::NONE ||
+                            prepareReturnStatus == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
                             prepareReturnStatus == ErrorStatus::MISSED_DEADLINE_PERSISTENT);
                 break;
             case DeadlineBoundType::UNLIMITED:
@@ -140,13 +148,19 @@
                 // of the switch statement.
                 EXPECT_EQ(ErrorStatus::NONE, prepareReturnStatus);
                 break;
+            case DeadlineBoundType::SHORT:
+                // Either the driver successfully completed the task in time or
+                // it aborted within the compliance time.
+                EXPECT_TRUE(prepareReturnStatus == ErrorStatus::NONE ||
+                            prepareReturnStatus == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
+                            prepareReturnStatus == ErrorStatus::MISSED_DEADLINE_PERSISTENT);
+                break;
         }
     }
     ASSERT_EQ(prepareReturnStatus == ErrorStatus::NONE, preparedModel.get() != nullptr);
 }
 
-void runPrepareModelTests(const sp<IDevice>& device, const Model& model,
-                          bool supportsPrepareModelDeadline) {
+void runPrepareModelTests(const sp<IDevice>& device, const Model& model) {
     // test priority
     for (auto priority : hidl_enum_range<Priority>{}) {
         SCOPED_TRACE("priority: " + toString(priority));
@@ -155,19 +169,17 @@
     }
 
     // test deadline
-    if (supportsPrepareModelDeadline) {
-        for (auto deadlineBound : deadlineBounds) {
-            SCOPED_TRACE("deadlineBound: " + toString(deadlineBound));
-            runPrepareModelTest(device, model, kDefaultPriority, deadlineBound);
-        }
+    for (auto deadlineBound : deadlineBounds) {
+        SCOPED_TRACE("deadlineBound: " + toString(deadlineBound));
+        runPrepareModelTest(device, model, kDefaultPriority, deadlineBound);
     }
 }
 
 static MaybeResults executeAsynchronously(const sp<IPreparedModel>& preparedModel,
-                                          const Request& request, DeadlineBoundType deadlineBound) {
+                                          const Request& request,
+                                          const OptionalTimePoint& deadline) {
     SCOPED_TRACE("asynchronous");
     const MeasureTiming measure = MeasureTiming::NO;
-    const OptionalTimePoint deadline = makeOptionalTimePoint(deadlineBound);
 
     // launch execution
     const sp<ExecutionCallback> callback = new ExecutionCallback();
@@ -187,10 +199,10 @@
 }
 
 static MaybeResults executeSynchronously(const sp<IPreparedModel>& preparedModel,
-                                         const Request& request, DeadlineBoundType deadlineBound) {
+                                         const Request& request,
+                                         const OptionalTimePoint& deadline) {
     SCOPED_TRACE("synchronous");
     const MeasureTiming measure = MeasureTiming::NO;
-    const OptionalTimePoint deadline = makeOptionalTimePoint(deadlineBound);
 
     // configure results callback
     MaybeResults results;
@@ -209,9 +221,10 @@
 void runExecutionTest(const sp<IPreparedModel>& preparedModel, const TestModel& testModel,
                       const Request& request, bool synchronous, DeadlineBoundType deadlineBound) {
     const ExecutionFunction execute = synchronous ? executeSynchronously : executeAsynchronously;
+    const auto deadline = makeDeadline(deadlineBound);
 
     // Perform execution and unpack results.
-    const auto results = execute(preparedModel, request, deadlineBound);
+    const auto results = execute(preparedModel, request, deadline);
     if (!results.has_value()) return;
     const auto& [status, outputShapes, timing] = results.value();
 
@@ -226,7 +239,8 @@
             // deadline has already passed when the driver would launch the
             // execution. In this case, the driver must return
             // MISSED_DEADLINE_*.
-            ASSERT_TRUE(status == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
+            ASSERT_TRUE(status == ErrorStatus::NONE ||
+                        status == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
                         status == ErrorStatus::MISSED_DEADLINE_PERSISTENT);
             return;
         case DeadlineBoundType::UNLIMITED:
@@ -235,6 +249,13 @@
             // of the switch statement.
             ASSERT_EQ(ErrorStatus::NONE, status);
             break;
+        case DeadlineBoundType::SHORT:
+            // Either the driver successfully completed the task in time or
+            // it aborted within the compliance time.
+            EXPECT_TRUE(status == ErrorStatus::NONE ||
+                        status == ErrorStatus::MISSED_DEADLINE_TRANSIENT ||
+                        status == ErrorStatus::MISSED_DEADLINE_PERSISTENT);
+            break;
     }
 
     // If the model output operands are fully specified, outputShapes must be either
@@ -268,32 +289,27 @@
     }
 }
 
-void runTests(const sp<IDevice>& device, const TestModel& testModel,
-              std::pair<bool, bool> supportsDeadlines) {
+void runTests(const sp<IDevice>& device, const TestModel& testModel) {
     // setup
-    const auto [supportsPrepareModelDeadline, supportsExecutionDeadline] = supportsDeadlines;
-    if (!supportsPrepareModelDeadline && !supportsExecutionDeadline) return;
     const Model model = createModel(testModel);
 
     // run prepare model tests
-    runPrepareModelTests(device, model, supportsPrepareModelDeadline);
+    runPrepareModelTests(device, model);
 
-    if (supportsExecutionDeadline) {
-        // prepare model
-        sp<IPreparedModel> preparedModel;
-        createPreparedModel(device, model, &preparedModel);
-        if (preparedModel == nullptr) return;
+    // prepare model
+    sp<IPreparedModel> preparedModel;
+    createPreparedModel(device, model, &preparedModel);
+    if (preparedModel == nullptr) return;
 
-        // run execution tests
-        const Request request = nn::convertToV1_3(createRequest(testModel));
-        runExecutionTests(preparedModel, testModel, request);
-    }
+    // run execution tests
+    const Request request = nn::convertToV1_3(createRequest(testModel));
+    runExecutionTests(preparedModel, testModel, request);
 }
 
 class DeadlineTest : public GeneratedTestBase {};
 
 TEST_P(DeadlineTest, Test) {
-    runTests(kDevice, kTestModel, mSupportsDeadlines);
+    runTests(kDevice, kTestModel);
 }
 
 INSTANTIATE_GENERATED_TEST(DeadlineTest,
diff --git a/neuralnetworks/1.3/vts/functional/ValidateModel.cpp b/neuralnetworks/1.3/vts/functional/ValidateModel.cpp
index 09e9922..8f2d4b7 100644
--- a/neuralnetworks/1.3/vts/functional/ValidateModel.cpp
+++ b/neuralnetworks/1.3/vts/functional/ValidateModel.cpp
@@ -44,18 +44,12 @@
 }
 
 static void validatePrepareModel(const sp<IDevice>& device, const std::string& message,
-                                 const Model& model, ExecutionPreference preference,
-                                 bool testDeadline) {
+                                 const Model& model, ExecutionPreference preference) {
     SCOPED_TRACE(message + " [prepareModel_1_3]");
 
-    OptionalTimePoint deadline;
-    if (testDeadline) {
-        deadline.nanosecondsSinceEpoch(std::numeric_limits<uint64_t>::max());
-    }
-
     sp<PreparedModelCallback> preparedModelCallback = new PreparedModelCallback();
     Return<ErrorStatus> prepareLaunchStatus = device->prepareModel_1_3(
-            model, preference, kDefaultPriority, deadline, hidl_vec<hidl_handle>(),
+            model, preference, kDefaultPriority, {}, hidl_vec<hidl_handle>(),
             hidl_vec<hidl_handle>(), HidlToken(), preparedModelCallback);
     ASSERT_TRUE(prepareLaunchStatus.isOk());
     ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, static_cast<ErrorStatus>(prepareLaunchStatus));
@@ -79,13 +73,12 @@
 // to the model does not leave this function.
 static void validate(const sp<IDevice>& device, const std::string& message, Model model,
                      const std::function<void(Model*)>& mutation,
-                     ExecutionPreference preference = ExecutionPreference::FAST_SINGLE_ANSWER,
-                     bool testDeadline = false) {
+                     ExecutionPreference preference = ExecutionPreference::FAST_SINGLE_ANSWER) {
     mutation(&model);
-    if (validExecutionPreference(preference) && !testDeadline) {
+    if (validExecutionPreference(preference)) {
         validateGetSupportedOperations(device, message, model);
     }
-    validatePrepareModel(device, message, model, preference, testDeadline);
+    validatePrepareModel(device, message, model, preference);
 }
 
 static uint32_t addOperand(Model* model) {
@@ -744,19 +737,9 @@
     }
 }
 
-///////////////////////// DEADLINE /////////////////////////
-
-static void deadlineTest(const sp<IDevice>& device, const Model& model) {
-    const std::string message = "deadlineTest: deadline not supported";
-    const auto noop = [](Model*) {};
-    validate(device, message, model, noop, ExecutionPreference::FAST_SINGLE_ANSWER,
-             /*testDeadline=*/true);
-}
-
 ////////////////////////// ENTRY POINT //////////////////////////////
 
-void validateModel(const sp<IDevice>& device, const Model& model,
-                   bool prepareModelDeadlineSupported) {
+void validateModel(const sp<IDevice>& device, const Model& model) {
     mutateOperandTypeTest(device, model);
     mutateOperandRankTest(device, model);
     mutateOperandScaleTest(device, model);
@@ -772,9 +755,6 @@
     addOperationInputTest(device, model);
     addOperationOutputTest(device, model);
     mutateExecutionPreferenceTest(device, model);
-    if (!prepareModelDeadlineSupported) {
-        deadlineTest(device, model);
-    }
 }
 
 }  // namespace android::hardware::neuralnetworks::V1_3::vts::functional
diff --git a/neuralnetworks/1.3/vts/functional/ValidateRequest.cpp b/neuralnetworks/1.3/vts/functional/ValidateRequest.cpp
index 20f4fe2..5e806e5 100644
--- a/neuralnetworks/1.3/vts/functional/ValidateRequest.cpp
+++ b/neuralnetworks/1.3/vts/functional/ValidateRequest.cpp
@@ -45,8 +45,7 @@
 // that use the request. Note that the request here is passed by value, and any
 // mutation to the request does not leave this function.
 static void validate(const sp<IPreparedModel>& preparedModel, const std::string& message,
-                     Request request, const std::function<void(Request*)>& mutation,
-                     bool testDeadline = false) {
+                     Request request, const std::function<void(Request*)>& mutation) {
     mutation(&request);
 
     // We'd like to test both with timing requested and without timing
@@ -59,18 +58,13 @@
     };
     MeasureTiming measure = (hash & 1) ? MeasureTiming::YES : MeasureTiming::NO;
 
-    OptionalTimePoint deadline;
-    if (testDeadline) {
-        deadline.nanosecondsSinceEpoch(std::numeric_limits<uint64_t>::max());
-    }
-
     // asynchronous
     {
         SCOPED_TRACE(message + " [execute_1_3]");
 
         sp<ExecutionCallback> executionCallback = new ExecutionCallback();
         Return<ErrorStatus> executeLaunchStatus =
-                preparedModel->execute_1_3(request, measure, deadline, {}, executionCallback);
+                preparedModel->execute_1_3(request, measure, {}, {}, executionCallback);
         ASSERT_TRUE(executeLaunchStatus.isOk());
         ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, static_cast<ErrorStatus>(executeLaunchStatus));
 
@@ -88,7 +82,7 @@
         SCOPED_TRACE(message + " [executeSynchronously_1_3]");
 
         Return<void> executeStatus = preparedModel->executeSynchronously_1_3(
-                request, measure, deadline, {},
+                request, measure, {}, {},
                 [](ErrorStatus error, const hidl_vec<OutputShape>& outputShapes,
                    const Timing& timing) {
                     ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, error);
@@ -100,7 +94,7 @@
 
     // burst
     // TODO(butlermichael): Check if we need to test burst in V1_3 if the interface remains V1_2.
-    if (!testDeadline) {
+    {
         SCOPED_TRACE(message + " [burst]");
 
         ASSERT_TRUE(nn::compliantWithV1_0(request));
@@ -143,7 +137,7 @@
     {
         SCOPED_TRACE(message + " [executeFenced]");
         Return<void> ret =
-                preparedModel->executeFenced(request, {}, MeasureTiming::NO, deadline, {}, {},
+                preparedModel->executeFenced(request, {}, MeasureTiming::NO, {}, {}, {},
                                              [](ErrorStatus error, const hidl_handle& handle,
                                                 const sp<IFencedExecutionCallback>& callback) {
                                                  ASSERT_EQ(ErrorStatus::INVALID_ARGUMENT, error);
@@ -174,23 +168,11 @@
     }
 }
 
-///////////////////////// DEADLINE ////////////////////////////////////
-
-static void deadlineTest(const sp<IPreparedModel>& preparedModel, const Request& request) {
-    const std::string message = "deadlineTest: deadline not supported";
-    const auto noop = [](Request*) {};
-    validate(preparedModel, message, request, noop, /*testDeadline=*/true);
-}
-
 ///////////////////////////// ENTRY POINT //////////////////////////////////
 
-void validateRequest(const sp<IPreparedModel>& preparedModel, const Request& request,
-                     bool executionDeadlineSupported) {
+void validateRequest(const sp<IPreparedModel>& preparedModel, const Request& request) {
     removeInputTest(preparedModel, request);
     removeOutputTest(preparedModel, request);
-    if (!executionDeadlineSupported) {
-        deadlineTest(preparedModel, request);
-    }
 }
 
 void validateRequestFailure(const sp<IPreparedModel>& preparedModel, const Request& request) {
diff --git a/neuralnetworks/1.3/vts/functional/VtsHalNeuralnetworks.cpp b/neuralnetworks/1.3/vts/functional/VtsHalNeuralnetworks.cpp
index 16341da..5b07034 100644
--- a/neuralnetworks/1.3/vts/functional/VtsHalNeuralnetworks.cpp
+++ b/neuralnetworks/1.3/vts/functional/VtsHalNeuralnetworks.cpp
@@ -123,11 +123,9 @@
 INSTANTIATE_DEVICE_TEST(NeuralnetworksHidlTest);
 
 // Forward declaration from ValidateModel.cpp
-void validateModel(const sp<IDevice>& device, const Model& model,
-                   bool prepareModelDeadlineSupported);
+void validateModel(const sp<IDevice>& device, const Model& model);
 // Forward declaration from ValidateRequest.cpp
-void validateRequest(const sp<IPreparedModel>& preparedModel, const Request& request,
-                     bool executionDeadlineSupported);
+void validateRequest(const sp<IPreparedModel>& preparedModel, const Request& request);
 // Forward declaration from ValidateRequest.cpp
 void validateRequestFailure(const sp<IPreparedModel>& preparedModel, const Request& request);
 // Forward declaration from ValidateBurst.cpp
@@ -147,17 +145,15 @@
     ASSERT_TRUE(ret_null.isOk());
 }
 
-void validateEverything(const sp<IDevice>& device, const Model& model, const Request& request,
-                        std::pair<bool, bool> supportsDeadlines) {
-    const auto [prepareModelDeadlineSupported, executionDeadlineSupported] = supportsDeadlines;
-    validateModel(device, model, prepareModelDeadlineSupported);
+void validateEverything(const sp<IDevice>& device, const Model& model, const Request& request) {
+    validateModel(device, model);
 
     // Create IPreparedModel.
     sp<IPreparedModel> preparedModel;
     createPreparedModel(device, model, &preparedModel);
     if (preparedModel == nullptr) return;
 
-    validateRequest(preparedModel, request, executionDeadlineSupported);
+    validateRequest(preparedModel, request);
     validateExecuteFenced(preparedModel, request);
 
     // TODO(butlermichael): Check if we need to test burst in V1_3 if the interface remains V1_2.
@@ -166,12 +162,10 @@
     validateBurst(preparedModel, request10);
 }
 
-void validateFailure(const sp<IDevice>& device, const Model& model, const Request& request,
-                     std::pair<bool, bool> supportsDeadlines) {
-    const bool prepareModelDeadlineSupported = supportsDeadlines.first;
+void validateFailure(const sp<IDevice>& device, const Model& model, const Request& request) {
     // TODO: Should this always succeed?
     //       What if the invalid input is part of the model (i.e., a parameter).
-    validateModel(device, model, prepareModelDeadlineSupported);
+    validateModel(device, model);
 
     // Create IPreparedModel.
     sp<IPreparedModel> preparedModel;
@@ -185,9 +179,9 @@
     const Model model = createModel(kTestModel);
     const Request request = nn::convertToV1_3(createRequest(kTestModel));
     if (kTestModel.expectFailure) {
-        validateFailure(kDevice, model, request, mSupportsDeadlines);
+        validateFailure(kDevice, model, request);
     } else {
-        validateEverything(kDevice, model, request, mSupportsDeadlines);
+        validateEverything(kDevice, model, request);
     }
 }