diff --git a/config/config.yaml b/config/config.yaml
index a9c764c56..6a1fd597f 100644
--- a/config/config.yaml
+++ b/config/config.yaml
@@ -117,4 +117,6 @@ RPM: 10
 ### repair operation on the content extracted from LLM's raw output. Warning, it improves the result but not fix all cases.
 # REPAIR_LLM_OUTPUT: false
 
-# PROMPT_FORMAT: json #json or markdown
\ No newline at end of file
+# PROMPT_FORMAT: json #json or markdown
+
+# DISABLE_LLM_PROVIDER_CHECK: false
diff --git a/metagpt/config.py b/metagpt/config.py
index 208b4fd7b..16df19a4c 100644
--- a/metagpt/config.py
+++ b/metagpt/config.py
@@ -107,7 +107,9 @@ class Config(metaclass=Singleton):
         self.gemini_api_key = self._get("GEMINI_API_KEY")
         self.ollama_api_base = self._get("OLLAMA_API_BASE")
         self.ollama_api_model = self._get("OLLAMA_API_MODEL")
-        _ = self.get_default_llm_provider_enum()
+
+        if not self._get("DISABLE_LLM_PROVIDER_CHECK"):
+            _ = self.get_default_llm_provider_enum()
 
         self.openai_base_url = self._get("OPENAI_BASE_URL")
         self.openai_proxy = self._get("OPENAI_PROXY") or self.global_proxy
diff --git a/metagpt/llm.py b/metagpt/llm.py
index 8763642f0..f1cb98dae 100644
--- a/metagpt/llm.py
+++ b/metagpt/llm.py
@@ -6,6 +6,8 @@
 @File    : llm.py
 """
 
+from typing import Optional
+
 from metagpt.config import CONFIG, LLMProviderEnum
 from metagpt.provider.base_gpt_api import BaseGPTAPI
 from metagpt.provider.human_provider import HumanProvider
@@ -14,6 +16,9 @@ from metagpt.provider.llm_provider_registry import LLM_REGISTRY
 _ = HumanProvider()  # Avoid pre-commit error
 
 
-def LLM(provider: LLMProviderEnum = CONFIG.get_default_llm_provider_enum()) -> BaseGPTAPI:
+def LLM(provider: Optional[LLMProviderEnum] = None) -> BaseGPTAPI:
     """get the default llm provider"""
+    if provider is None:
+        provider = CONFIG.get_default_llm_provider_enum()
+
     return LLM_REGISTRY.get_provider(provider)