diff --git a/api/ai/ai_response_model.py b/api/ai/ai_response_model.py
index 4beab39..0d9ef33 100644
--- a/api/ai/ai_response_model.py
+++ b/api/ai/ai_response_model.py
@@ -55,4 +55,11 @@ class ResponseMetadata(BaseModel):
 class StreamResponse(BaseModel):
     results: List[WorkflowResult]
     metadata: ResponseMetadata
-    errors: List[Any]
\ No newline at end of file
+    errors: List[Any]
+
+class EnhanceRequest(BaseModel):
+    prompt: str
+    model: str = "claude-sonnet-4-20250514"
+
+class EnhanceResponse(BaseModel):
+    enhanced_prompt: str
\ No newline at end of file
diff --git a/api/ai/ai_service.py b/api/ai/ai_service.py
index 1bd7b65..1875de4 100644
--- a/api/ai/ai_service.py
+++ b/api/ai/ai_service.py
@@ -12,11 +12,13 @@
     WorkflowResult, 
     ResponseMetadata,
     AvailableModelsResponse,
-    ModelInfo
+    ModelInfo,
+    EnhanceResponse
 )
 from api.db.pg_database import SessionLocal
 from api.llm.router import get_model_router
 from api.external_api import get_related_segment_ids, get_segment_content
+from api.ai.prompts import ENHANCE_META_PROMPT
 from fastapi import HTTPException
 
 
@@ -154,4 +156,19 @@ def get_available_models_service() -> AvailableModelsResponse:
             context_window=model_data.get("context_window")
         )
     
-    return AvailableModelsResponse(models=models)
\ No newline at end of file
+    return AvailableModelsResponse(models=models)
+
+async def enhance_prompt_service(prompt: str, model: str = "claude-sonnet-4-20250514") -> EnhanceResponse:
+    model_router = get_model_router()
+    print(model)
+    if not model_router.validate_model_availability(model):
+        raise HTTPException(
+            status_code=503,
+            detail=f"Model '{model}' is not available. Check API key configuration."
+        )
+
+    llm = model_router.get_model(model, temperature=0.7, max_tokens=4096)
+    response = await llm.ainvoke(ENHANCE_META_PROMPT.format(prompt=prompt))
+    enhanced_text = response.content if hasattr(response, "content") else str(response)
+
+    return EnhanceResponse(enhanced_prompt=enhanced_text)
\ No newline at end of file
diff --git a/api/ai/ai_view.py b/api/ai/ai_view.py
index 29af25a..201d675 100644
--- a/api/ai/ai_view.py
+++ b/api/ai/ai_view.py
@@ -1,5 +1,5 @@
-from api.ai.ai_response_model import StreamRequest, AvailableModelsResponse
-from api.ai.ai_service import run_workflow_service, stream_workflow_service, get_available_models_service
+from api.ai.ai_response_model import StreamRequest, AvailableModelsResponse, EnhanceRequest, EnhanceResponse
+from api.ai.ai_service import run_workflow_service, stream_workflow_service, get_available_models_service, enhance_prompt_service
 from fastapi import APIRouter, HTTPException, Depends
 from fastapi.responses import StreamingResponse
 from starlette import status
@@ -82,3 +82,15 @@ async def stream_workflow(
         media_type="text/event-stream",
         headers=SSE_HEADERS
     )
+
+@ai_router.post("/enhance", status_code=status.HTTP_200_OK, response_model=EnhanceResponse)
+async def enhance_prompt(
+    payload: EnhanceRequest,
+    authentication_credential: Annotated[
+        HTTPAuthorizationCredentials, Depends(oauth2_scheme)
+    ],
+):
+    return await enhance_prompt_service(
+        prompt=payload.prompt,
+        model=payload.model,
+    )
\ No newline at end of file
diff --git a/api/ai/prompts.py b/api/ai/prompts.py
index 05396fe..6bad3a2 100644
--- a/api/ai/prompts.py
+++ b/api/ai/prompts.py
@@ -37,3 +37,19 @@ def get_specialized_prompt(
         additional_context_section=additional_context_section,
         source_text_block=source_text_block,
     )
+
+
+ENHANCE_META_PROMPT = """You are an expert prompt engineer. Your task is to enhance and improve the given system prompt while preserving the user's original intent and purpose.
+
+Guidelines for enhancement:
+- Make the instructions clearer and more specific
+- Add structure (e.g., role definition, constraints, output format) where beneficial
+- Remove ambiguity and vagueness
+- Maintain the original language and tone the user intended
+- Keep it concise — don't add unnecessary verbosity
+- If the prompt references specific domains or tasks, sharpen those references
+
+Return ONLY the enhanced system prompt text. Do not include any explanation, commentary, or markdown formatting around it.
+
+Original system prompt to enhance:
+{prompt}"""
\ No newline at end of file