Update app.py

app.py

@@ -6,8 +6,8 @@ import time
 from html import escape
 
 # Model config
-PRIMARY_MODEL = "Smilyai-labs/Sam-reason-A1"
-FALLBACK_MODEL = "Smilyai-labs/Sam-reason-S2.1"
+PRIMARY_MODEL = "Smilyai-labs/Sam-reason-A3"
+FALLBACK_MODEL = "Smilyai-labs/Sam-reason-A1"
 USAGE_LIMIT = 5
 RESET_AFTER_SECONDS = 20 * 60  # 20 minutes
 device = "cuda" if torch.cuda.is_available() else "cpu"
@@ -85,16 +85,16 @@ def respond(message, history, reasoning_enabled, request: gr.Request):
     for output in generate_stream(prompt, use_fallback=use_fallback):
         formatted = format_thinking(output)
         history[-1][1] = f"{formatted}<br><sub style='color:gray'>({model_used})</sub>"
-        yield history, history, f"🧠 A1 messages left: {remaining}"
+        yield history, history, f"🧠 A3 messages left: {remaining}"
 
 def clear_chat():
-    return [], [], "🧠 A1 messages left: 10"
+    return [], [], "🧠 A3 messages left: 5"
 
 # UI
 with gr.Blocks() as demo:
     gr.Markdown("# 🤖 SamAI – Reasoning Chat")
     model_status = gr.Textbox(interactive=False, label="Model Status")
-    usage_counter = gr.Textbox(value="🧠 A1 messages left: 10", interactive=False, show_label=False)
+    usage_counter = gr.Textbox(value="🧠 A3 messages left: 10", interactive=False, show_label=False)
     chat_box = gr.Chatbot(type="tuples")
     chat_state = gr.State([])