app.py

import gradio as gr
import google.generativeai as genai
from PIL import Image
import io
import numpy as np

def generate_content(api_key, prompt, image_data):

    genai.configure(api_key=api_key)  

    # Convert uploaded binary data into a PIL image.
    # image = Image.open(io.BytesIO(image_data))
    image = Image.fromarray(np.uint8(image_data)).convert('RGB')

    model = genai.GenerativeModel('gemini-pro-vision')

    response = model.generate_content([prompt, image])
    response.resolve()
    return response.text



with gr.Blocks() as app:
    gr.Markdown("# gemini-pro-vision demo")
    gr.Markdown("This is a simple example of image understanding, supporting the upload of only one image,Enter your API key to start the conversation. Get a free API key from here: https://ai.google.dev/")
    api_key_input = gr.Textbox(label="API Key", type="password", placeholder="Enter your API Key", lines=1)

    # image_input = gr.File(label="Upload an image", type="binary")
    image_input = gr.Image(label="Upload an image")
    prompt_input = gr.Textbox(label="Prompt", placeholder="Enter your Prompt", value="Describe this picture",lines=3)
    generate_button = gr.Button("Generate Content")
    output = gr.Markdown()
    
    generate_button.click(
        generate_content, 
        inputs=[api_key_input, prompt_input, image_input],
        outputs=output
    )

app.launch()