Text, Image, and Audio Inputs

Gemini accepts a list of content parts. This list can mix text and binary data.

Image Input (PIL)

For images, efficient handling is key.

import PIL.Image

# Load local image
img = PIL.Image.open('cat.jpg')

# Pass directly
response = model.generate_content(["Describe this image:", img])

Audio/Video Input (File API)

For larger files like video/audio, you cannot just pass raw bytes in the list efficiently. You upload them first.

# Upload
video_file = genai.upload_file(path="webinar.mp4")

# Wait for processing (Video takes time to ingest)
import time
while video_file.state.name == "PROCESSING":
    time.sleep(2)
    video_file = genai.get_file(video_file.name)

# Generate
response = model.generate_content([video_file, "Summarize the speakers points."])

Summary

• Small Images -> PIL objects.
• Large Audio/Video -> upload_file() API.

In the next lesson, we discuss Combining Modalities.