From 31dcc44c2030b285ea939638d8fda2bdfcf2e318 Mon Sep 17 00:00:00 2001
From: Debanjum <debanjum@gmail.com>
Date: Sun, 18 May 2025 14:40:57 -0700
Subject: [PATCH] Output tokens >> reasoning tokens to avoid early response
 termination.

---
 src/khoj/processor/conversation/google/utils.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/src/khoj/processor/conversation/google/utils.py b/src/khoj/processor/conversation/google/utils.py
index c527bf72..22cd8d0f 100644
--- a/src/khoj/processor/conversation/google/utils.py
+++ b/src/khoj/processor/conversation/google/utils.py
@@ -36,7 +36,9 @@ logger = logging.getLogger(__name__)
 
 gemini_clients: Dict[str, genai.Client] = {}
 
-MAX_OUTPUT_TOKENS_GEMINI = 8192
+# Output tokens should be more than reasoning tokens.
+# This avoids premature response termination.
+MAX_OUTPUT_TOKENS_GEMINI = 20000
 MAX_REASONING_TOKENS_GEMINI = 10000
 
 SAFETY_SETTINGS = [