[Edit]

llm.py

@@ -5,18 +5,19 @@ from dotenv import load_dotenv
 
 load_dotenv()
 
-api_key = os.getenv("GOOGLE_API_KEY")
-if not api_key:
-    raise ValueError("GOOGLE_API_KEY environment variable is not set. Please add it to your .env file")
+# Support multiple Gemini keys (comma-separated or single key)
+api_keys = os.getenv("GOOGLE_API_KEYS") or os.getenv("GOOGLE_API_KEY")
+if not api_keys:
+    raise ValueError("No Gemini API keys found in GOOGLE_API_KEYS or GOOGLE_API_KEY environment variable.")
 
-print(f"Google API Key loaded: {api_key[:10]}..." if api_key else "No API key found")
-genai.configure(api_key=api_key)
+api_keys = [k.strip() for k in api_keys.split(",") if k.strip()]
+print(f"Loaded {len(api_keys)} Gemini API key(s)")
 
 def query_gemini(questions, contexts):
-    try:
-        context = "\n\n".join(contexts)
-        questions_text = "\n".join([f"{i+1}. {q}" for i, q in enumerate(questions)])
-        prompt = f"""
+    context = "\n\n".join(contexts)
+    questions_text = "\n".join([f"{i+1}. {q}" for i, q in enumerate(questions)])
+
+    prompt = f"""
 You are an expert insurance assistant generating formal yet user-facing answers to policy questions and Other Human Questions. Your goal is to write professional, structured answers that reflect the language of policy documents — but are still human-readable and easy to understand.
 
 🧠 FORMAT & TONE GUIDELINES:
@@ -73,19 +74,37 @@ Respond with only the following JSON — no explanations, no comments, no markdo
 
 Your task: For each question, provide a complete, professional, and clearly written answer in 2–3 sentences using a formal but readable tone.
 """
-        model = genai.GenerativeModel('gemini-2.5-flash-lite')
-        response = model.generate_content(prompt)
-        response_text = response.text.strip()
+
+    last_exception = None
+
+    for key_idx, key in enumerate(api_keys):
         try:
+            genai.configure(api_key=key)
+            model = genai.GenerativeModel("gemini-2.5-flash-lite")
+            response = model.generate_content(prompt)
+            response_text = response.text.strip()
+
+            # Clean JSON from wrapped code blocks
             if response_text.startswith("```json"):
                 response_text = response_text.replace("```json", "").replace("```", "").strip()
             elif response_text.startswith("```"):
                 response_text = response_text.replace("```", "").strip()
-            parsed_response = json.loads(response_text)
-            return parsed_response
-        except json.JSONDecodeError:
-            print(f"Failed to parse JSON response: {response_text}")
-            return {"answers": ["Error parsing response"] * len(questions)}
-    except Exception as e:
-        print(f"Error in query_gemini: {str(e)}")
-        return {"answers": [f"Error generating response: {str(e)}"] * len(questions)}
+
+            parsed = json.loads(response_text)
+            if "answers" in parsed and isinstance(parsed["answers"], list):
+                return parsed
+            else:
+                raise ValueError("Invalid response format received from Gemini.")
+
+        except Exception as e:
+            msg = str(e).lower()
+            if "429" in msg or "quota" in msg or "rate limit" in msg or "exceeded" in msg:
+                print(f"Gemini key {key[:8]}... quota/rate limited. Trying next key ({key_idx+1}/{len(api_keys)}).")
+                last_exception = e
+                continue
+            else:
+                print(f"Error using Gemini key {key[:8]}...: {e}")
+                return {"answers": [f"Error generating response: {str(e)}"] * len(questions)}
+
+    print(f"All Gemini API keys failed. Last error: {last_exception}")
+    return {"answers": [f"Error generating response: {str(last_exception)}"] * len(questions)}