mjtaherikia's picture
New model used for image tool.
e1031ca
import requests
from google import genai
from google.genai import types
from smolagents import tool
from modules.prompts import image_description_prompt
@tool
def image_description(question: str, image_url: str) -> str:
"""
Recieves an image url and analyze it first, then return a what is needed based on prompt.
Args:
question: a question about given image
image_url: the url of image which should be considered as a reference for answering the question.
"""
image_bytes = requests.get(image_url).content
image = types.Part.from_bytes(
data=image_bytes, mime_type="image/jpeg"
)
client = genai.Client()
response = client.models.generate_content(
model="gemini-2.5-flash-preview-05-20",
contents=[image_description_prompt.format(question=question), image],
)
return response.text