How do I read image data from a URL in Python?

Tips for heavy files: streaming

When dealing with large images, your memory could blow up like a balloon 🎈. Utilize streaming to avoid this:

response = requests.get('YOUR_IMAGE_URL', stream=True) # Drunken master: staggers to avoid blackout.
response.raw.decode_content = True
image = Image.open(response.raw)


Enabling stream=True sets requests to download in chunks, instead of one big haul. Do remember to use response.raw.decode_content = True to correctly decode the content stream!

Uncommon image formats: the rare beast taming guide

Think you are dealing with a rarely-seen image format? No worries, tame it by converting:

import requests
from io import BytesIO
from PIL import Image

response = requests.get('YOUR_IMAGE_URL')
image = Image.open(BytesIO(response.content))

if image.format == 'SOME_RARE_FORMAT':    # You rarely seen unicorn? Tame it!
    image = image.convert('RGB')  # Convert to common format, if needed


The image.format attribute tells the tale of its origin. You can convert the rare beast to a commoner with image.convert() .

Syntax of invincibility: error handling

Error handling: the most important spell in Programmers vs Bugs universe. Let's see how to make it to the final round!

import requests
from PIL import Image
from io import BytesIO

try:
    response = requests.get('YOUR_IMAGE_URL')
    response.raise_for_status()  
    image = Image.open(BytesIO(response.content))
except requests.HTTPError as http_err:
    print(f'HTTP error occurred: {http_err}') # Uh-oh, the 4 horsemen arrive!
except Exception as err:
    print(f'Other error occurred: {err}') # If it's not horsemen, it's zombies. 💀 


The response.raise_for_status() lifts the veil on HTTP errors- revealing the 4 horsemen! The Exception catchall is the last line of defense against unknown forces! Who's next - zombies?!

Faster than light: optimizing speed and memory

You want to be flash and save the day without burning out memory. Here's the superpower kit:

• Large files: Use response.iter_content(chunk_size=1024) to bite off what you can chew.
• Caching: Like squirrels 🐿️, stash your fetched images to reduce server load and quicken future retrievals.
• Concurrency: Clone yourself á la Dr.Stange with Python's ThreadPoolExecutor for download multiverse exploration!

Jupyter/IPython printers: direct display

For the folks living in the Jupyter universe, here's your special weapon for direct image display:

from IPython.display import Image, display

display(Image(url='YOUR_IMAGE_URL'))  # Just-Jupyter-Things: One wand wave, image appears!


The display function coupled with Image from IPython.display, you can wave your wand and have the image rendered in your notebook!