from io import BytesIO
from agno.agent import Agent, RunOutput # noqa
from agno.media import Image
from agno.models.google import Gemini
from PIL import Image as PILImage
# No system message should be provided (Gemini requires only the image)
agent = Agent(
model=Gemini(
id="gemini-2.0-flash-exp-image-generation",
response_modalities=["Text", "Image"],
)
)
# Print the response in the terminal
response = agent.run(
"Can you add a Llama in the background of this image?",
images=[Image(filepath="tmp/test_photo.png")],
)
# Retrieve and display generated images using get_last_run_output
run_response = agent.get_last_run_output()
if run_response and isinstance(run_response, RunOutput) and run_response.images:
for image_response in run_response.images:
image_bytes = image_response.content
if image_bytes:
image = PILImage.open(BytesIO(image_bytes))
image.show()
# Save the image to a file
# image.save("generated_image.png")
else:
print("No images found in run response")
Create a virtual environment
Terminal
and create a python virtual environment.python3 -m venv .venv
source .venv/bin/activate
Prepare your image
tmp/test_photo.png
or update the filepath in the code to point to your image.Set your API key
export GOOGLE_API_KEY=xxx
Install libraries
pip install -U google-genai pillow agno
Run Agent
python cookbook/models/google/gemini/image_editing.py