Update app.py

app.py

@@ -3,6 +3,7 @@ import requests
 import json
 
 OLLAMA_URL = "http://localhost:11434"
+HF_API_URL = "https://api-inference.huggingface.co/models/openai/whisper-large-v3-turbo"
 
 MODELS = {
     "Qwen2.5-Coder 1.5B (Fastest)": "qwen2.5-coder:1.5b",
@@ -17,6 +18,24 @@ def check_ollama():
     except:
         return False
 
+def transcribe_audio(audio):
+    if audio is None:
+        return ""
+    
+    try:
+        with open(audio, "rb") as f:
+            data = f.read()
+        
+        response = requests.post(HF_API_URL, data=data, timeout=60)
+        
+        if response.status_code == 200:
+            result = response.json()
+            return result.get("text", "")
+        else:
+            return f"[STT Error: {response.status_code}]"
+    except Exception as e:
+        return f"[STT Error: {e}]"
+
 def chat_stream(message, history, model_name, temperature, max_tokens):
     if not check_ollama():
         yield "⏳ Ollama starting... wait 30 seconds and try again."
@@ -131,9 +150,12 @@ with gr.Blocks(title="Axon v5.1", theme=gr.themes.Soft(primary_hue="purple")) as
         with gr.TabItem("💬 Chat"):
             chatbot = gr.Chatbot(height=400)
             with gr.Row():
-                msg = gr.Textbox(placeholder="Ask about coding...", show_label=False, scale=9)
+                msg = gr.Textbox(placeholder="Ask about coding...", show_label=False, scale=7)
+                audio_input = gr.Audio(sources=["microphone"], type="filepath", label="🎤", scale=2)
                 send = gr.Button("Send", variant="primary", scale=1)
-            clear = gr.Button("Clear")
+            with gr.Row():
+                clear = gr.Button("Clear")
+                transcribe_btn = gr.Button("🎤 Transcribe", variant="secondary")
             gr.Examples(["Write a Python quicksort function", "Explain async/await in JavaScript"], inputs=msg)
         
         with gr.TabItem("⚡ Generate"):
@@ -166,6 +188,7 @@ with gr.Blocks(title="Axon v5.1", theme=gr.themes.Soft(primary_hue="purple")) as
     msg.submit(respond, [msg, chatbot, model_dropdown, temperature, max_tokens], [chatbot, msg])
     send.click(respond, [msg, chatbot, model_dropdown, temperature, max_tokens], [chatbot, msg])
     clear.click(lambda: [], None, chatbot)
+    transcribe_btn.click(transcribe_audio, audio_input, msg)
     gen_btn.click(generate_code, [gen_prompt, gen_lang, model_dropdown, max_tokens], gen_output)
     explain_btn.click(explain_code, [explain_input, model_dropdown, max_tokens], explain_output)
     fix_btn.click(fix_code, [fix_input, fix_error, model_dropdown, max_tokens], fix_output)