diff --git a/pilot/model/llm_out/gpt4all_llm.py b/pilot/model/llm_out/gpt4all_llm.py
index 5ea72f911..7a39a8012 100644
--- a/pilot/model/llm_out/gpt4all_llm.py
+++ b/pilot/model/llm_out/gpt4all_llm.py
@@ -19,5 +19,5 @@ def gpt4all_generate_stream(model, tokenizer, params, device, max_position_embed
 
     while t.is_alive():
         yield sys.stdout.output
-        time.sleep(0.1)
+        time.sleep(0.01)
     t.join()