diff --git a/common/arg.cpp b/common/arg.cpp
index 733947930..1e20d7464 100644
--- a/common/arg.cpp
+++ b/common/arg.cpp
@@ -1787,7 +1787,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
     ).set_examples({LLAMA_EXAMPLE_SERVER}).set_env("LLAMA_ARG_CACHE_REUSE"));
     add_opt(common_arg(
         {"--standby-timeout"}, "N",
-        string_format("time that must pass since a request has been served, before the server stops automatically (default: %d)", params.standby_timeout),
+        string_format("seconds that must pass since a request has been served, before the server stops automatically (default: %d)", params.standby_timeout),
         [](common_params & params, int value) {
             params.standby_timeout = value;
         }
diff --git a/common/common.h b/common/common.h
index b1f19eb32..9f0582365 100644
--- a/common/common.h
+++ b/common/common.h
@@ -306,7 +306,7 @@ struct common_params {
     int32_t timeout_write  = timeout_read; // http write timeout in seconds
     int32_t n_threads_http = -1;           // number of threads to process HTTP requests (TODO: support threadpool)
     int32_t n_cache_reuse  = 0;            // min chunk size to reuse from the cache via KV shifting
-    int32_t standby_timeout  = 0;          // time that must pass since a request has been processed before server terminates in order to save resources. If -1, then never terminate automatically.
+    int32_t standby_timeout  = 0;          // seconds that must pass since a request has been processed before server terminates in order to save resources. If -1, then never terminate automatically.
 
     std::string hostname      = "127.0.0.1";
     std::string public_path   = "";                                                                         // NOLINT