diff --git a/src/plugins/intel_gpu/include/intel_gpu/plugin/common_utils.hpp b/src/plugins/intel_gpu/include/intel_gpu/plugin/common_utils.hpp
index e3ec998df78..df2caf5bcfc 100644
--- a/src/plugins/intel_gpu/include/intel_gpu/plugin/common_utils.hpp
+++ b/src/plugins/intel_gpu/include/intel_gpu/plugin/common_utils.hpp
@@ -121,6 +121,7 @@ inline void ForceExit() {
 void convert_and_copy(const ov::ITensor* src, cldnn::memory::ptr dst, cldnn::stream& stream);
 void convert_and_copy(const cldnn::memory::ptr src, ov::ITensor const* dst, const cldnn::stream& stream);
 void convert_and_copy(const ov::ITensor* src, ov::ITensor const* dst, const cldnn::stream& stream);
+void convert_and_copy(const cldnn::memory::ptr src, cldnn::memory::ptr dst, cldnn::stream& stream);
 
 }  // namespace intel_gpu
 
diff --git a/src/plugins/intel_gpu/src/plugin/common_utils.cpp b/src/plugins/intel_gpu/src/plugin/common_utils.cpp
index 7dfa876570b..9ab7b74b82d 100644
--- a/src/plugins/intel_gpu/src/plugin/common_utils.cpp
+++ b/src/plugins/intel_gpu/src/plugin/common_utils.cpp
@@ -138,6 +138,20 @@ void convert_and_copy(const cldnn::memory::ptr src, ov::ITensor const* dst, cons
     return ::convert_and_copy(src_ptr, src_et, dst_ptr, dst_et, size, src->get_layout());
 }
 
+void convert_and_copy(const cldnn::memory::ptr src, cldnn::memory::ptr dst, cldnn::stream& stream) {
+    const bool blocking = true;
+    auto src_et = src->get_layout().data_type;
+    auto dst_et = dst->get_layout().data_type;
+
+    cldnn::mem_lock<uint8_t, cldnn::mem_lock_type::read> src_lock(src, stream);
+    const void* src_ptr = src_lock.data();
+
+    size_t size = ov::shape_size(src->get_layout().get_shape());
+    ov::Tensor tmp_tensor(dst_et, src->get_layout().get_shape());
+    ::convert_and_copy(src_ptr, src_et, tmp_tensor.data(), dst_et, size, src->get_layout());
+    dst->copy_from(stream, tmp_tensor.data(), blocking);
+}
+
 void convert_and_copy(const ov::ITensor* src, ov::ITensor const* dst, const cldnn::stream& stream) {
     auto src_et = src->get_element_type();
     auto dst_et = dst->get_element_type();
diff --git a/src/plugins/intel_gpu/src/plugin/sync_infer_request.cpp b/src/plugins/intel_gpu/src/plugin/sync_infer_request.cpp
index 51861988f1f..8b21669f2d8 100644
--- a/src/plugins/intel_gpu/src/plugin/sync_infer_request.cpp
+++ b/src/plugins/intel_gpu/src/plugin/sync_infer_request.cpp
@@ -675,7 +675,13 @@ std::vector<cldnn::event::ptr> SyncInferRequest::prepare_input(const std::string
     bool convert_needed = is_convert_required(element_type, device_tensor_et);
 
     if (is_remote) {
-        m_plugin_inputs[name] = user_tensor_wrapper;
+        if (convert_needed) {
+            m_plugin_inputs[name] = { create_device_tensor(pshape,
+                                                           cldnn::element_type_to_data_type(element_type),
+                                                           false), TensorOwner::PLUGIN };
+        } else {
+            m_plugin_inputs[name] = user_tensor_wrapper;
+        }
     } else if (is_usm_host_tensor && !convert_needed && can_use_usm_host(engine)) {
         if (element_type != cldnn::element_type_to_data_type(element_type)) {
             m_plugin_inputs[name] = {std::make_shared<RemoteTensorImpl>(m_context,
@@ -753,14 +759,17 @@ std::vector<cldnn::event::ptr> SyncInferRequest::prepare_input(const std::string
     }
 
     cldnn::event::ptr ret_event = nullptr;
-    if (!is_remote) {
-        if (convert_needed) {
-            convert_and_copy(user_tensor.get(), device_tensor.get(), stream);
+    if (!is_remote && !convert_needed) {
+        auto src_ptr = static_cast<uint8_t*>(user_tensor->data());
+        if (!same_host_mem(memory, src_ptr)) {
+            ret_event = memory->copy_from(stream, src_ptr, false);
+        }
+    }
+    if (convert_needed) {
+        if (is_remote) {
+            convert_and_copy(remote_ptr->get_memory(), device_tensor->get_memory(), stream);
         } else {
-            auto src_ptr = static_cast<uint8_t*>(user_tensor->data());
-            if (!same_host_mem(memory, src_ptr)) {
-                ret_event = memory->copy_from(stream, src_ptr, false);
-            }
+            convert_and_copy(user_tensor.get(), device_tensor.get(), stream);
         }
     }