Agent with Image Input

AWS Bedrock supports image input with models like amazon.nova-pro-v1:0. You can use this to analyze images and get information about them.

Code

cookbook/models/aws/bedrock/image_agent.py

from pathlib import Path
from agno.agent import Agent
from agno.media import Image
from agno.models.aws import AwsBedrock
from agno.tools.duckduckgo import DuckDuckGoTools

agent = Agent(
    model=AwsBedrock(id="amazon.nova-pro-v1:0"),
    tools=[DuckDuckGoTools()],
    markdown=True,
)

image_path = Path(__file__).parent.joinpath("sample.jpg")

# Read the image file content as bytes
with open(image_path, "rb") as img_file:
    image_bytes = img_file.read()

agent.print_response(
    "Tell me about this image and give me the latest news about it.",
    images=[
        Image(content=image_bytes, format="jpeg"),
    ],
)

Usage

1. Create a virtual environment

   Open the Terminal and create a python virtual environment.

   Mac

   python3 -m venv .venv
   source .venv/bin/activate
   
2. Set your AWS Credentials

   export AWS_ACCESS_KEY_ID=***
   export AWS_SECRET_ACCESS_KEY=***
   export AWS_REGION=***
   
3. Install libraries

   pip install -U boto3 duckduckgo-search agno
   
4. Add an Image

   Place an image file named sample.jpg in the same directory as your script.

5. Run Agent

   Mac

   python cookbook/models/aws/bedrock/image_agent.py