diff --git a/backend/python/openvoice/test.py b/backend/python/openvoice/test.py
index 262917b3..82f08785 100644
--- a/backend/python/openvoice/test.py
+++ b/backend/python/openvoice/test.py
@@ -19,7 +19,7 @@ class TestBackendServicer(unittest.TestCase):
         This method sets up the gRPC service by starting the server
         """
         self.service = subprocess.Popen(["python3", "backend.py", "--addr", "localhost:50051"])
-        time.sleep(10)
+        time.sleep(30)
 
     def tearDown(self) -> None:
         """
diff --git a/backend/python/vllm/install.sh b/backend/python/vllm/install.sh
index 022cf8bf..9078b81b 100755
--- a/backend/python/vllm/install.sh
+++ b/backend/python/vllm/install.sh
@@ -13,7 +13,9 @@ if [ "x${BUILD_PROFILE}" == "xintel" ]; then
     EXTRA_PIP_INSTALL_FLAGS+=" --upgrade --index-strategy=unsafe-first-match"
 fi
 
-if [ "x${BUILD_TYPE}" == "x" ]; then
+# We don't embed this into the images as it is a large dependency and not always needed.
+# Besides, the speed inference are not actually usable in the current state for production use-cases.
+if [ "x${BUILD_TYPE}" == "x" ] && [ "x${FROM_SOURCE}" == "xtrue" ]; then
         ensureVenv
         # https://docs.vllm.ai/en/v0.6.1/getting_started/cpu-installation.html
         if [ ! -d vllm ]; then