Update app.py

app.py

@@ -7,17 +7,18 @@ import base64
 # Configure the Gemini API client with the hardcoded API key
 GOOGLE_API_KEY = "AIzaSyDL5Rilo7ptJpUOZdY6wy8PJYUcVcnDADs"
 client = genai.Client(api_key=GOOGLE_API_KEY)
+GEMINI_MODEL_NAME = 'gemini-2.5-flash-image-preview'
 
 def process_image(image, prompt):
     try:
-        # Convert Gradio image (PIL Image) to base64
-        buffered = BytesIO()
-        image.save(buffered, format="PNG")
-        img_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")
-
         # Prepare the content for the Gemini API
-        contents = [
-            {
+        contents = []
+        if image:
+            # Convert Gradio image (PIL Image) to base64
+            buffered = BytesIO()
+            image.save(buffered, format="PNG")
+            img_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")
+            contents.append({
                 "parts": [
                     {"text": prompt},
                     {
@@ -27,24 +28,26 @@ def process_image(image, prompt):
                         }
                     }
                 ]
-            }
-        ]
+            })
+        else:
+            # Text-to-image generation
+            contents.append({"parts": [{"text": prompt}]})
 
         # Call the Gemini API
         response = client.models.generate_content(
-            model="gemini-2.5-flash-image-preview",
+            model=GEMINI_MODEL_NAME,
             contents=contents
         )
 
         # Process the response
-        for part in response.candidates[0].content.parts:
-            if part.inline_data is not None:
-                # Decode the generated image
-                img_data = base64.b64decode(part.inline_data.data)
-                img = Image.open(BytesIO(img_data))
-                return img
-            elif part.text is not None:
-                return f"Text response: {part.text}"
+        for candidate in response.candidates:
+            for part in candidate.content.parts:
+                if hasattr(part, 'inline_data') and part.inline_data:
+                    # Decode the generated image
+                    img_data = base64.b64decode(part.inline_data.data)
+                    return Image.open(BytesIO(img_data))
+                elif part.text:
+                    return f"Text response: {part.text}"
         
         return "No image or text returned by the model."
     
@@ -52,23 +55,35 @@ def process_image(image, prompt):
         return f"Error: {str(e)}"
 
 # Create the Gradio interface
-title = "Gemini 2.5 Flash Image Editor (Nano Banana)"
-description = "Upload an image and provide a text prompt to generate or edit images using Google's Gemini 2.5 Flash Image Preview model."
-interface = gr.Interface(
-    fn=process_image,
-    inputs=[
-        gr.Image(type="pil", label="Upload Image"),
-        gr.Textbox(label="Prompt", placeholder="e.g., 'Add a nano-banana to the image in a fancy restaurant setting'")
-    ],
-    outputs=gr.Image(type="pil", label="Generated Image"),
-    title=title,
-    description=description,
-    examples=[
-        [None, "Create a minimalist composition with a single red maple leaf in the bottom-right, soft lighting, square image"],
-        [None, "Generate an image of a cat eating a nano-banana in a fancy restaurant"]
-    ]
-)
+css = '''
+.grid-container img {object-fit: contain}
+.grid-container {display: grid; grid-template-columns: 1fr}
+'''
+
+with gr.Blocks(theme=gr.themes.Citrus(), css=css) as demo:
+    gr.HTML('''
+    <img src='https://huggingface.co/spaces/multimodalart/nano-banana/resolve/main/nano_banana_pros_light.png' style='margin: 0 auto; max-width: 500px' />
+    <h3 style='text-align:center'>Nano Banana: Gemini 2.5 Flash Image Preview</h3>
+    ''')
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            image_input = gr.Image(type="pil", label="Upload Image (Optional)", file_types=["image"])
+            prompt_input = gr.Textbox(
+                label="Prompt",
+                placeholder="e.g., 'Add a nano-banana to the image in a fancy restaurant setting' or 'Generate a cat eating a nano-banana'"
+            )
+            generate_button = gr.Button("Generate", variant="primary")
+
+        with gr.Column(scale=1):
+            output_image = gr.Image(label="Generated Image", type="pil")
+
+    # Event handler
+    generate_button.click(
+        fn=process_image,
+        inputs=[image_input, prompt_input],
+        outputs=[output_image]
+    )
 
-# Launch the app
 if __name__ == "__main__":
-    interface.launch()
+    demo.launch()