Skip using max_tokens as input to the extract questions step, as that's not used for max_output

src/khoj/processor/conversation/anthropic/anthropic_chat.py

@@ -25,7 +25,6 @@ def extract_questions_anthropic(
     conversation_log={},
     api_key=None,
     temperature=0,
-    max_tokens=100,
     location_data: LocationData = None,
 ):
     """
@@ -71,7 +70,6 @@ def extract_questions_anthropic(
         model_name=model,
         temperature=temperature,
         api_key=api_key,
-        max_tokens=max_tokens,
     )
 
     # Extract, Clean Message from Claude's Response

src/khoj/routers/api.py

@@ -353,7 +353,6 @@ async def extract_references_and_questions(
                 api_key=api_key,
                 conversation_log=meta_log,
                 location_data=location_data,
-                max_tokens=conversation_config.max_prompt_size,
             )
 
     # Collate search results as context for GPT