Feature Proposal: Support Passing Images as Base64 in `agent.print_response` [ Official Standard Way ]

[MANISH007700]

Description

Currently, the method add_images_to_message_content supports adding images to message content through URLs.

• While the method already processes these inputs effectively, there is room for improvement to better support the use of base64-encoded images, Read PIL Image and also Byte Image.

The goal is to make it easier to directly pass base64-encoded image strings as input without requiring prior manual validation or transformations. This would streamline workflows for users who already work with base64 images in their pipelines.

Proposed Changes Update the agent.print_response functionality to explicitly accept and process base64-encoded image strings. Modify the logic in add_images_to_message_content to prioritize and validate base64 strings more robustly.

PREVIEW CODE

agent.print_response(
    [
        {"type": "text", "text": "What's in this image, describe in 1 sentence"},
        {
            "type": "image_url",
            "image_url": f"data:image/jpeg;base64,{base64_string}",
        },
    ]
)

Lmk if you want me to pick this up and will get it done.

cc : @manthanguptaa @ashpreetbedi

[ysolanky]

@MANISH007700 that is a great catch! Please feel free to take this up. I look forward to reviewing your PR! Thank you!

[MANISH007700]

@ysolanky can you check this issue first - Viz Model Hallucination I'll push the new code after this issue is resolved.

Thanks