app.py

import gradio as gr
import google.generativeai as genai
from PIL import Image
from io import BytesIO
import base64

# List of popular styles
STYLES = [
    "Photorealistic", "Oil Painting", "Watercolor", "Anime", 
    "Studio Ghibli", "Black and White", "Polaroid", "Sketch", 
    "3D Render", "Pixel Art", "Cyberpunk", "Steampunk", 
    "Art Nouveau", "Pop Art", "Minimalist"
]

# Default negative prompt
DEFAULT_NEGATIVE_PROMPT = """
ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, 
extra limbs, disfigured, deformed, body out of frame, bad anatomy, watermark, signature, 
cut off, low contrast, underexposed, overexposed, bad art, beginner, amateur, distorted face
"""

def enhance_prompt(model, prompt, style):
    enhanced_prompt_request = f"Enhance the following prompt for image generation in the style of {style}. Make it more detailed and vivid, while keeping the original intent: '{prompt}'"
    
    response = model.generate_content(enhanced_prompt_request)
    
    return response.text

def generate_image(model, enhanced_prompt, style, negative_prompt):
    full_prompt = f"{enhanced_prompt}\nStyle: {style}\nNegative prompt: {negative_prompt}"
    
    response = model.generate_content([full_prompt, "generate an image"])
    
    if response.parts[0].image:
        image_bytes = response.parts[0].image.to_bytes()
        image = Image.open(BytesIO(image_bytes))
        
        # Convert PIL Image to base64 string
        buffered = BytesIO()
        image.save(buffered, format="PNG")
        img_str = base64.b64encode(buffered.getvalue()).decode()
        
        return f"data:image/png;base64,{img_str}"
    else:
        return None

def process_and_generate(api_key, prompt, style, negative_prompt):
    genai.configure(api_key=api_key)
    text_model = genai.GenerativeModel('gemini-2.0-flash-lite')
    image_model = genai.GenerativeModel('imagen-3.0-generate-002')
    enhanced_prompt = enhance_prompt(text_model, prompt, style)
    image_url = generate_image(image_model, enhanced_prompt, style, negative_prompt)
    return image_url, enhanced_prompt

with gr.Blocks() as demo:
    gr.Markdown("# Google Gemini Image Generator")
    
    with gr.Row():
        with gr.Column(scale=1):
            api_key = gr.Textbox(label="Google AI API Key", type="password")
            prompt = gr.Textbox(label="Prompt")
            style = gr.Dropdown(label="Style", choices=STYLES)
            negative_prompt = gr.Textbox(label="Negative Prompt", value=DEFAULT_NEGATIVE_PROMPT)
            generate_btn = gr.Button("Generate Image")
        
        with gr.Column(scale=1):
            image_output = gr.Image(label="Generated Image")
            enhanced_prompt_output = gr.Textbox(label="Enhanced Prompt")
    
    generate_btn.click(
        process_and_generate,
        inputs=[api_key, prompt, style, negative_prompt],
        outputs=[image_output, enhanced_prompt_output]
    )

demo.launch()