diff --git a/meshai/cli/configurator.py b/meshai/cli/configurator.py
index 3e5dabe..050d4e8 100644
--- a/meshai/cli/configurator.py
+++ b/meshai/cli/configurator.py
@@ -256,6 +256,8 @@ class Configurator:
             table.add_row("3", "Base URL", self.config.llm.base_url)
             table.add_row("4", "Model", self.config.llm.model)
             table.add_row("5", "System Prompt", f"[dim]{len(self.config.llm.system_prompt)} chars[/dim]")
+            use_prompt = getattr(self.config.llm, 'use_system_prompt', True)
+            table.add_row("6", "Use System Prompt", self._status_icon(use_prompt))
             table.add_row("0", "Back", "")
 
             console.print(table)
@@ -300,6 +302,10 @@ class Configurator:
                     if value != self.config.llm.system_prompt:
                         self.config.llm.system_prompt = value
                         self.modified = True
+            elif choice == 6:
+                current = getattr(self.config.llm, 'use_system_prompt', True)
+                self.config.llm.use_system_prompt = not current
+                self.modified = True
 
     def _weather_settings(self) -> None:
         """Weather settings submenu."""
diff --git a/meshai/config.py b/meshai/config.py
index 1c123c6..199d321 100644
--- a/meshai/config.py
+++ b/meshai/config.py
@@ -211,6 +211,7 @@ class LLMConfig:
         "Keep responses VERY brief - under 250 characters total. "
         "Be concise but friendly. No markdown formatting."
     )
+    use_system_prompt: bool = True  # Toggle to disable sending system prompt
 
     # Fallback settings
     fallback: Optional[LLMBackendConfig] = None
diff --git a/meshai/router.py b/meshai/router.py
index 2d4911a..dba56e9 100644
--- a/meshai/router.py
+++ b/meshai/router.py
@@ -130,10 +130,15 @@ class MessageRouter:
         history = await self.history.get_history_for_llm(message.sender_id)
 
         # Generate response with user_id for memory optimization
+        # Use system prompt only if enabled in config
+        system_prompt = ""
+        if getattr(self.config.llm, 'use_system_prompt', True):
+            system_prompt = self.config.llm.system_prompt
+
         try:
             response = await self.llm.generate(
                 messages=history,
-                system_prompt=self.config.llm.system_prompt,
+                system_prompt=system_prompt,
                 max_tokens=300,
                 user_id=message.sender_id,  # Enable memory optimization
             )