diff --git a/ai/litellm-config.yaml b/ai/litellm-config.yaml
index b3a8c51..325e2b0 100644
--- a/ai/litellm-config.yaml
+++ b/ai/litellm-config.yaml
@@ -41,7 +41,6 @@ model_list:
       stream_timeout: 600
       supports_system_messages: true # Llama supports system messages
       stream: true # Enable streaming by default
-      max_tokens: 4096 # Cap completion tokens to leave room for input
 
   # Embeddings - BGE Large (Port 8002)
   - model_name: bge-large-en