return model used

clients/groq_client.py

@@ -178,25 +178,25 @@ class GroqClient:
 
 # Non-streaming ask method for backwards compatibility
 # ============================================
-    def ask(self, raw_query: str) -> str:
+    def ask(self, raw_query: str) -> tuple[str, str | None]:
         if (
             not raw_query
             or raw_query is None
             or raw_query == ""
             or len(raw_query) > 1000
         ):
-            return "Please provide a valid query under 1,000 characters."
+            return "Please provide a valid query under 1,000 characters.", None
 
         if raw_query.lower() in GREETINGS_TRIGGERS:
-            return random.choice(GREETINGS)
+            return random.choice(GREETINGS), None
 
         try:
             docs = self.retriever.invoke(raw_query)
         except Exception as e:
-            return f"Error retrieving documents: {e}"
+            return f"Error retrieving documents: {e}", None
 
         if not docs:
-            return self.handle_unknown_query()
+            return self.handle_unknown_query(), None
 
         context = "\n".join([d.page_content for d in docs])
         fallback = self.handle_unknown_query()
@@ -242,7 +242,7 @@ class GroqClient:
                 )
                 response = completion.choices[0].message.content
                 if response and response.strip():
-                    return response.strip()
+                    return response.strip(), model
                 else:
                     continue  # Try next model
 
@@ -253,10 +253,10 @@ class GroqClient:
                     continue
                 else:
                     # For other errors, return immediately
-                    return f"Error while calling LLM: {e}"
+                    return f"Error while calling LLM: {e}", None
 
         # If all models fail
-        return "I'm temporarily experiencing high demand. Please try again in a few minutes or rephrase your question."
+        return "I'm temporarily experiencing high demand. Please try again in a few minutes or rephrase your question.", None
 
 # Streaming ask method (Generator)
 # ============================================

main.py

@@ -43,18 +43,20 @@ def get_groq_client():
 async def chat_endpoint(request: QueryRequest):
     try:
         groq_client = get_groq_client()
-        response = groq_client.ask(request.query)
+        response, model = groq_client.ask(request.query)
         if response.startswith("Error while"):
-            return {"status_code": 500, "response": response, "next_questions": []}
+            return {"status_code": 500, "response": response, "model": model, "next_questions": []}
         return {
             "status_code": 200,
             "response": response,
+            "model": model,
             "next_questions": groq_client.get_next_questions(request.ip),
         }
     except Exception as e:
         return {
             "status_code": 500,
             "response": f"Internal error: {e}",
+            "model": None,
             "next_questions": [],
         }