diff --git a/docs/IE_DG/supported_plugins/GNA.md b/docs/IE_DG/supported_plugins/GNA.md
index ca035194f61..7a4303c2d25 100644
--- a/docs/IE_DG/supported_plugins/GNA.md
+++ b/docs/IE_DG/supported_plugins/GNA.md
@@ -109,7 +109,7 @@ When specifying key values as raw strings (that is, when using Python API), omit
 
 | Parameter Name                    | Parameter Values                                          | Default Value     | Description                                                              |
 | :---------------------------------| :---------------------------------------------------------| :-----------| :------------------------------------------------------------------------|
-| `KEY_GNA_COMPACT_MODE`            | `YES`/`NO`                                                | `YES`       | Reuse I/O buffers to save space (makes debugging harder)                 |
+| `KEY_GNA_COMPACT_MODE`            | `YES`/`NO`                                                | `NO`       | Reuse I/O buffers to save space (makes debugging harder)                 |
 | `KEY_GNA_SCALE_FACTOR`            | `FP32` number                                             | 1.0         | Scale factor to use for input quantization                               |
 | `KEY_GNA_DEVICE_MODE`             | `GNA_AUTO`/`GNA_HW`/`GNA_SW_EXACT`/`GNA_SW_FP32` | `GNA_AUTO`  | One of the modes described <a name="execution-models">Execution Models</a> |
 | `KEY_GNA_FIRMWARE_MODEL_IMAGE`    | `std::string`                                             | `""`        | Name for embedded model binary dump file                                 |
diff --git a/inference-engine/src/gna_plugin/descriptions/gna_flags.hpp b/inference-engine/src/gna_plugin/descriptions/gna_flags.hpp
index 7536880fe1f..1f71da6f613 100644
--- a/inference-engine/src/gna_plugin/descriptions/gna_flags.hpp
+++ b/inference-engine/src/gna_plugin/descriptions/gna_flags.hpp
@@ -10,7 +10,7 @@ namespace GNAPluginNS {
 struct GNAFlags {
     uint8_t gna_lib_async_threads_num = 1;
 
-    bool compact_mode = true;
+    bool compact_mode = false;
     bool exclusive_async_requests = false;
     bool uniformPwlDesign = false;
     bool gna_openmp_multithreading = false;
diff --git a/inference-engine/tests/unit/gna/gna_plugin_config_test.cpp b/inference-engine/tests/unit/gna/gna_plugin_config_test.cpp
index c930b59f187..d07f4d63282 100644
--- a/inference-engine/tests/unit/gna/gna_plugin_config_test.cpp
+++ b/inference-engine/tests/unit/gna/gna_plugin_config_test.cpp
@@ -16,7 +16,7 @@ const std::map<std::string, std::string>  supportedConfigKeysWithDefaults = {
     {GNA_CONFIG_KEY(FIRMWARE_MODEL_IMAGE), ""},
     {GNA_CONFIG_KEY(FIRMWARE_MODEL_IMAGE_GENERATION), ""},
     {GNA_CONFIG_KEY(DEVICE_MODE), GNAConfigParams::GNA_SW_EXACT},
-    {GNA_CONFIG_KEY(COMPACT_MODE), CONFIG_VALUE(YES)},
+    {GNA_CONFIG_KEY(COMPACT_MODE), CONFIG_VALUE(NO)},
     {CONFIG_KEY(EXCLUSIVE_ASYNC_REQUESTS), CONFIG_VALUE(NO)},
     {GNA_CONFIG_KEY(PRECISION), Precision(Precision::I16).name()},
     {GNA_CONFIG_KEY(PWL_UNIFORM_DESIGN), CONFIG_VALUE(NO)},