Update app.py

app.py

@@ -1,48 +1,74 @@
-import os
 import gradio as gr
 from google import genai
-from google.genai import types
+from PIL import Image
+from io import BytesIO
+import base64
 
-# Set your Google API key
-os.environ["GOOGLE_API_KEY"] = "AIzaSyDL5Rilo7ptJpUOZdY6wy8PJYUcVcnDADs"
+# Configure the Gemini API client with the hardcoded API key
+GOOGLE_API_KEY = "AIzaSyDL5Rilo7ptJpUOZdY6wy8PJYUcVcnDADs"
+client = genai.Client(api_key=GOOGLE_API_KEY)
 
-# Initialize the Google Generative AI client
-client = genai.Client(api_key=os.environ["GOOGLE_API_KEY"])
+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")
 
-MODEL_NAME = "gemini-2.5-flash-image-preview"
+        # Prepare the content for the Gemini API
+        contents = [
+            {
+                "parts": [
+                    {"text": prompt},
+                    {
+                        "inline_data": {
+                            "mime_type": "image/png",
+                            "data": img_base64
+                        }
+                    }
+                ]
+            }
+        ]
 
-def remix_images(image1, image2, prompt):
-    try:
-        # Run request to Google Generative AI
-        response = client.models.generate_images(
-            model=MODEL_NAME,
-            prompt=prompt,
-            images=[image1, image2]
+        # Call the Gemini API
+        response = client.models.generate_content(
+            model="gemini-2.5-flash-image-preview",
+            contents=contents
         )
 
-        # Extract the first generated image
-        if response.generated_images and len(response.generated_images) > 0:
-            img = response.generated_images[0]
-            return img.image  # returns PIL.Image
-        else:
-            return None
+        # 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}"
+        
+        return "No image or text returned by the model."
+    
     except Exception as e:
         return f"Error: {str(e)}"
 
-# Gradio UI
-with gr.Blocks() as demo:
-    gr.Markdown("## 🖼️ Google Gemini 2.5 Flash Image Preview (via Gradio)")
-
-    with gr.Row():
-        img1 = gr.Image(type="filepath", label="Upload Image 1")
-        img2 = gr.Image(type="filepath", label="Upload Image 2")
-
-    prompt = gr.Textbox(label="Prompt", placeholder="Describe how to remix the images...")
-
-    btn = gr.Button("Generate")
-    output = gr.Image(label="Generated Image")
-
-    btn.click(fn=remix_images, inputs=[img1, img2, prompt], outputs=output)
+# 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"]
+    ]
+)
 
+# Launch the app
 if __name__ == "__main__":
-    demo.launch()
+    interface.launch()