Update app.py

app.py

@@ -8,15 +8,30 @@ def to_markdown(text):
   text = text.replace('•', ' *')
   return textwrap.indent(text, '> ', lambda line: True)
 
-def chat(message, history):
-  """Generates a response to the user's message using the Gemini API."""
+def chat(message, history, img=None):
+  """Generates a response to the user's message, optionally using an image."""
   genai.configure(api_key='AIzaSyCMBk81YmILNTok8hd6tYtJaevp1qbl6I0')  # Replace with your actual API key
-  model = genai.GenerativeModel('gemini-pro')
+
+  text_model = genai.GenerativeModel('gemini-pro')
+  vision_model = genai.GenerativeModel('gemini-pro-vision')
 
   try:
-    response = model.generate_content(message, stream=True)
+    if img is not None:
+      # Process image with vision model
+      image_response = vision_model.generate_content(img)
+      image_response.resolve()  # Wait for response completion
+      image_text = to_markdown(image_response.text)
+
+      # Combine image and text for unified response
+      prompt = f"{message}\n{image_text}"
+      response = text_model.generate_content(prompt, stream=True)
+    else:
+      # Use only text model
+      response = text_model.generate_content(message, stream=True)
+
     for chunk in response:
       return to_markdown(chunk.text)  # Format as Markdown
+
   except Exception as e:
     print(f"Error during generation: {e}")
     return "An error occurred while generating the response. Please try again later."
@@ -25,7 +40,8 @@ chat_interface = gr.ChatInterface(
   fn=chat,
   title="Gemini Chat",
   description="Chat with an AI assistant powered by Gemini",
-  theme="soft"
+  theme="soft",
+  inputs=[gr.Textbox(lines=1, label="Enter your message"), gr.Image(optional=True)],
 )
 
 chat_interface.launch()