Automated deployment via API

app.py

@@ -361,8 +361,8 @@ from huggingface_hub import InferenceClient
 import os
 
 # Initialize Inference Client
-# Using Phi-3 Mini: High availability, low latency, works well on free tier
-client = InferenceClient("microsoft/Phi-3-mini-4k-instruct", token=os.environ.get("HF_TOKEN"))
+# Using Zephyr-7B Beta: The gold standard for free-tier Spaces hosting
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta", token=os.environ.get("HF_TOKEN"))
 
 def generate_ai_response(query):
     """Generate AI response using direct InferenceClient"""