diff --git a/rag/llm/__init__.py b/rag/llm/__init__.py
index eccf3f7df..14e9a8a19 100644
--- a/rag/llm/__init__.py
+++ b/rag/llm/__init__.py
@@ -43,6 +43,7 @@ FACTORY_DEFAULT_BASE_URL = {
     SupportedLiteLLMProvider.Tongyi_Qianwen: "https://dashscope.aliyuncs.com/compatible-mode/v1",
     SupportedLiteLLMProvider.Dashscope: "https://dashscope.aliyuncs.com/compatible-mode/v1",
     SupportedLiteLLMProvider.Moonshot: "https://api.moonshot.cn/v1",
+    SupportedLiteLLMProvider.Ollama: "",
 }
 
 
diff --git a/rag/llm/chat_model.py b/rag/llm/chat_model.py
index 1948ee848..5575aa390 100644
--- a/rag/llm/chat_model.py
+++ b/rag/llm/chat_model.py
@@ -1362,7 +1362,7 @@ class LiteLLMBase(ABC):
         self.prefix = LITELLM_PROVIDER_PREFIX.get(self.provider, "")
         self.model_name = f"{self.prefix}{model_name}"
         self.api_key = key
-        self.base_url = base_url or FACTORY_DEFAULT_BASE_URL.get(self.provider, "")
+        self.base_url = (base_url or FACTORY_DEFAULT_BASE_URL.get(self.provider, "")).rstrip('/')
         # Configure retry parameters
         self.max_retries = kwargs.get("max_retries", int(os.environ.get("LLM_MAX_RETRIES", 5)))
         self.base_delay = kwargs.get("retry_interval", float(os.environ.get("LLM_BASE_DELAY", 2.0)))