diff --git a/samples/cpp/benchmark_app/benchmark_app.hpp b/samples/cpp/benchmark_app/benchmark_app.hpp
index 413083e159a..64c6adcc8e5 100644
--- a/samples/cpp/benchmark_app/benchmark_app.hpp
+++ b/samples/cpp/benchmark_app/benchmark_app.hpp
@@ -272,7 +272,7 @@ DEFINE_uint32(nireq, 0, infer_requests_count_message);
 DEFINE_uint32(nthreads, 0, infer_num_threads_message);
 
 /// @brief Number of streams to use for inference on the CPU (also affects Hetero cases)
-DEFINE_string(nstreams, "AUTO", infer_num_streams_message);
+DEFINE_string(nstreams, "", infer_num_streams_message);
 
 /// @brief The percentile which will be reported in latency metric
 DEFINE_uint32(latency_percentile, 50, infer_latency_percentile_message);
diff --git a/src/inference/src/threading/ie_istreams_executor.cpp b/src/inference/src/threading/ie_istreams_executor.cpp
index 7bd8067809c..c39eeb2b630 100644
--- a/src/inference/src/threading/ie_istreams_executor.cpp
+++ b/src/inference/src/threading/ie_istreams_executor.cpp
@@ -122,7 +122,7 @@ void IStreamsExecutor::Config::SetConfig(const std::string& key, const std::stri
             OPENVINO_UNREACHABLE("Wrong value for property key ",
                                  ov::num_streams.name(),
                                  ". Expected non negative numbers (#streams) or ",
-                                 "ov::NumStreams::NUMA|ov::NumStreams::AUTO, Got: ",
+                                 "ov::NumStreams(ov::NumStreams::NUMA|ov::NumStreams::AUTO), Got: ",
                                  streams);
         }
     } else if (key == CONFIG_KEY(CPU_THREADS_NUM) || key == ov::inference_num_threads) {
diff --git a/src/plugins/intel_gpu/src/plugin/plugin.cpp b/src/plugins/intel_gpu/src/plugin/plugin.cpp
index 0c39decf82f..1a077012763 100644
--- a/src/plugins/intel_gpu/src/plugin/plugin.cpp
+++ b/src/plugins/intel_gpu/src/plugin/plugin.cpp
@@ -225,7 +225,7 @@ std::map<std::string, std::string> Plugin::ConvertPerfHintsToConfig(
                 config[ov::num_streams.name()] = std::to_string(1);
             } else if (mode_name == CONFIG_VALUE(THROUGHPUT)) {
                 config[PluginConfigParams::KEY_GPU_THROUGHPUT_STREAMS] = CONFIG_VALUE(GPU_THROUGHPUT_AUTO);
-                config[ov::num_streams.name()] = ov::Any::make<ov::NumStreams>(ov::NumStreams::AUTO).as<std::string>();
+                config[ov::num_streams.name()] = ov::util::to_string(ov::NumStreams(ov::NumStreams::AUTO));
                 //disabling the throttling temporarily to set the validation (that is switching to the hints) perf baseline
                 //checking throttling (to avoid overriding what user might explicitly set in the incoming config or previously via SetConfig)
                 // const auto bInConfig = config.find(GPUConfigParams::KEY_GPU_PLUGIN_THROTTLE) != config.end() ||
diff --git a/src/plugins/intel_myriad/common/src/configuration/options/ov_throughput_streams.cpp b/src/plugins/intel_myriad/common/src/configuration/options/ov_throughput_streams.cpp
index fa9f13e501f..60ebbe5ea87 100644
--- a/src/plugins/intel_myriad/common/src/configuration/options/ov_throughput_streams.cpp
+++ b/src/plugins/intel_myriad/common/src/configuration/options/ov_throughput_streams.cpp
@@ -46,7 +46,7 @@ details::Category OvThroughputStreamsOption::category() {
 
 std::string OvThroughputStreamsOption::defaultValue() {
     std::stringstream ss;
-    ss << ov::NumStreams::AUTO;
+    ss << ov::NumStreams(ov::NumStreams::AUTO);
     return ss.str();
 }