If you need to process data in memory without saving it to disk first (common in AWS Lambda), use download_fileobj() . This requires a writable object opened in .
For small files where you just want the raw content (like a JSON config or text snippet), get_object() is the standard tool. It returns a StreamingBody that you can read directly. python boto3 download s3 file
import boto3 import io s3 = boto3.client('s3') buffer = io.BytesIO() s3.download_fileobj('my-bucket', 'data/image.jpg', buffer) # Reset buffer position to read it buffer.seek(0) image_data = buffer.read() Use code with caution. If you need to process data in memory