Update main.py

main.py

@@ -38,7 +38,7 @@ LLM_MODEL = "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8"
 MAX_SOURCES_TO_PROCESS = 20  # Increased for more research
 MAX_CONCURRENT_REQUESTS = 2
 SEARCH_TIMEOUT = 300  # 5 minutes for longer research
-TOTAL_TIMEOUT = 360  # Adjusted accordingly
+TOTAL_TIMEOUT = 600  # Increased to allow more time for generation
 REQUEST_DELAY = 3.0
 RETRY_ATTEMPTS = 5
 RETRY_DELAY = 5.0
@@ -643,7 +643,7 @@ async def run_deep_research_stream(query: str, search_time: int = 300) -> AsyncG
                 "data": f"Synthesizing comprehensive report from {successful_sources} sources..."
             })
 
-            max_output_tokens = min(16000, int(time_remaining * 6))  # Increased for longer report
+            max_output_tokens = 16000  # Fixed to allow long response
 
             report_prompt = f"""Compose an in-depth analysis report on "{query}".