Update app.py

app.py

@@ -14,14 +14,16 @@ from smolagents import CodeAgent, DuckDuckGoSearchTool, tool, InferenceClientMod
 
 # 0. SHARED CONFIG
 HF_TOKEN = os.getenv("HF_TOKEN")
+# 7B is the sweet spot for free serverless inference in 2026
+MODEL_ID = "Qwen/Qwen2.5-7B-Instruct"
 
 # ==========================================
 # PART 1: LLAMAINDEX AGENT
 # ==========================================
 li_llm = HuggingFaceInferenceAPI(
-    model_name="Qwen/Qwen2.5-7B-Instruct",
+    model_name=MODEL_ID,
     token=HF_TOKEN,
-    provider="together"  # Routing through a stable provider
+    provider="together" 
 )
 
 def get_tokyo_time() -> str:
@@ -47,7 +49,7 @@ async def chat_llama(message, history):
 # PART 2: SMOLAGENTS
 # ==========================================
 smol_model = InferenceClientModel(
-    model_id="Qwen/Qwen2.5-Coder-32B-Instruct",
+    model_id=MODEL_ID,
     token=HF_TOKEN,
     provider="together"
 )
@@ -60,7 +62,6 @@ def weather_tool(location: str) -> str:
     """
     return f"The weather in {location} is currently sunny and 22°C."
 
-# Ensure ddgs is installed via requirements.txt for this to work
 smol_agent = CodeAgent(
     model=smol_model,
     tools=[weather_tool, DuckDuckGoSearchTool()]
@@ -74,10 +75,11 @@ def chat_smol(message, history):
         return f"Smolagents Error: {str(e)}"
 
 # ==========================================
-# PART 3: GRADIO UI
+# PART 3: UNIFIED GRADIO UI
 # ==========================================
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# 🤖 Dual-Agent Framework Comparison")
+    gr.Markdown("# 🤖 Consolidated AI Agent Space")
+    gr.Markdown(f"Currently using **{MODEL_ID}** via Together AI Provider.")
     
     with gr.Tab("LlamaIndex (Workflow)"):
         gr.ChatInterface(fn=chat_llama)