diff --git a/examples/server/server.cpp b/examples/server/server.cpp
index ecae8ecc3..9b653a2f6 100644
--- a/examples/server/server.cpp
+++ b/examples/server/server.cpp
@@ -756,8 +756,6 @@ bool parse_options_completion(json body, llama_server_context& llama, Response &
 
 int main(int argc, char **argv)
 {
-  llama_init_backend();
-
   // own arguments required by this example
   gpt_params params;
   server_params sparams;
@@ -775,6 +773,8 @@ int main(int argc, char **argv)
     params.model_alias = params.model;
   }
 
+  llama_init_backend();
+
   fprintf(stderr, "%s: build = %d (%s)\n", __func__, BUILD_NUMBER, BUILD_COMMIT);
   fprintf(stderr, "system_info: n_threads = %d / %d | %s\n\n", params.n_threads,
           std::thread::hardware_concurrency(), llama_print_system_info());