Python script that can downloads public and private profiles images and videos, like Gallery with photos or videos. It saves the data in the folder.

How it works:

  • Log in in instragram using selenium and navigate to the profile

  • Checking the availability of Instagram profile if it's private or existing

  • Creates a folder with the name of your choice

  • Gathering urls from images and videos

  • Using threads and multiprocessing improving the execution speed

My code:

import requests
import time
from selenium import webdriver
from selenium.common.exceptions import NoSuchElementException, TimeoutException
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By
from multiprocessing.dummy import Pool
from concurrent.futures import ThreadPoolExecutor
from typing import *
import argparse
import shutil
from functools import reduce

class PrivateException(Exception):

class InstagramPV:
    MAX_WORKERS: int = 8
    N_PROCESSES: int = 8

    BASE_URL = 'https://www.instagram.com/'
    PROFILE_URL_FMT = BASE_URL + '{name}/'
    LOGIN_URL = BASE_URL + 'accounts/login'

    def __init__(self, username: str, password: str, folder: Path, profile_name: str):

        :param username: Username or E-mail for Log-in in Instagram
        :param password: Password for Log-in in Instagram
        :param folder: Folder name that will save the posts
        :param profile_name: The profile name that will search
        self.username = username
        self.password = password
        self.folder = folder
        self.http_base = requests.Session()
        self.profile_name = profile_name
        self.links: List[str] = []
        self.pictures: List[str] = []
        self.videos: List[str] = []
        self.posts: int = 0
        self.driver = webdriver.Chrome()

    def __enter__(self):
        return self

    def __exit__(self, exc_type, exc_val, exc_tb):

    def check_availability(self) -> None:
        Checking Status code, Taking number of posts, Privacy and followed by viewer
        Raise Error if the Profile is private and not following by viewer
        :return: None
        search = self.http_base.get(self.PROFILE_URL_FMT.format(name=self.profile_name), params={'__a': 1})

        load_and_check = search.json()
        user = (
            load_and_check.get('graphql', {})
                .get('user', {})
        self.posts = (
                .get('edge_owner_to_timeline_media', {})

        privacy = (

        followed_by_viewer = (

        if privacy and not followed_by_viewer:
            raise PrivateException('[!] Account is private')

    def create_folder(self) -> None:
        """Create the folder name"""

    def login(self) -> None:
        """Login To Instagram"""
        WebDriverWait(self.driver, 10).until(EC.presence_of_element_located((By.TAG_NAME, 'form')))
        submit = self.driver.find_element_by_tag_name('form')

        """Check For Invalid Credentials"""
            var_error = WebDriverWait(self.driver, 4).until(EC.presence_of_element_located((By.CLASS_NAME, 'eiCW-')))
            raise ValueError(var_error.text)
        except TimeoutException:

            """Close Notifications"""
            notifications = WebDriverWait(self.driver, 20).until(
                EC.presence_of_element_located((By.XPATH, '//button[text()="Not Now"]')))
        except NoSuchElementException:

        """Taking cookies"""
        cookies = {
            cookie['name']: cookie['value']
            for cookie in self.driver.get_cookies()


        """Check for availability"""



    def posts_urls(self) -> None:
        """Taking the URLs from posts and appending in self.links"""
        elements = self.driver.find_elements_by_xpath('//a[@href]')
        for elem in elements:
            urls = elem.get_attribute('href')
            if urls not in self.links and 'p' in urls.split('/'):

    def scroll_down(self) -> None:
        """Scrolling down the page and taking the URLs"""
        last_height = 0
        while True:
            self.driver.execute_script('window.scrollTo(0, document.body.scrollHeight);')
            new_height = self.driver.execute_script("return document.body.scrollHeight")
            if new_height == last_height:
            last_height = new_height

    def submit_links(self) -> None:
        """Gathering Images and Videos and pass to function <fetch_url> Using ThreadPoolExecutor"""


        print('[!] Ready for video - images'.title())
        print(f'[*] extracting {len(self.links)} posts , please wait...'.title())

        with ThreadPoolExecutor(max_workers=self.MAX_WORKERS) as executor:
            for link in self.links:
                executor.submit(self.fetch_url, link)

    def fetch_url(self, url: str) -> None:
        This function extracts images and videos
        :param url: Taking the url
        :return None
        logging_page_id = self.http_base.get(url, params={'__a': 1}).json()

        if self.get_fields(logging_page_id, '__typename') == 'GraphImage':
            image_url = self.get_fields(logging_page_id, 'display_url')

        elif self.get_fields(logging_page_id, '__typename') == 'GraphVideo':
            video_url = self.get_fields(logging_page_id, 'video_url')

        elif self.get_fields(logging_page_id, '__typename') == 'GraphSidecar':
            for sidecar in self.get_fields(logging_page_id, 'edge_sidecar_to_children', 'edges'):
                if self.get_fields(sidecar, '__typename') == 'GraphImage':
                    image_url = self.get_fields(sidecar, 'display_url')
                    video_url = self.get_fields(sidecar, 'video_url')
            print(f'Warning {url}: has unknown type of {self.get_fields(logging_page_id,"__typename")}')

    def get_fields(nodes: Dict[str, Any], *keys: Iterable[str]) -> Any:
        :param nodes: The json data from the link using only the first two keys 'graphql' and 'shortcode_media'
        :param keys: Keys that will be add to the nodes and will have the results of 'type' or 'URL'
        :return: The value of the key <fields>
        media = ['graphql', 'shortcode_media', *keys]
        if list(nodes.keys())[0] == 'node':
            media = ['node', *keys]
        field = reduce(dict.get, media, nodes)
        return field

    def download_video(self, new_videos: Tuple[int, str]) -> None:
        Saving the video content
        :param new_videos: Tuple[int,str]
        :return: None
        number, link = new_videos

        with open(self.folder / f'Video{number}.mp4', 'wb') as f, \
                self.http_base.get(link, stream=True) as response:
            shutil.copyfileobj(response.raw, f)

    def images_download(self, new_pictures: Tuple[int, str]) -> None:
        Saving the picture content
        :param new_pictures: Tuple[int, str]
        :return: None

        number, link = new_pictures
        with open(self.folder / f'Image{number}.jpg', 'wb') as f, \
                self.http_base.get(link, stream=True) as response:
            shutil.copyfileobj(response.raw, f)

    def downloading_video_images(self) -> None:
        """Using multiprocessing for Saving Images and Videos"""
        print('[*] ready for saving images and videos!'.title())
        picture_data = enumerate(self.pictures)
        video_data = enumerate(self.videos)
        pool = Pool(self.N_PROCESSES)
        pool.map(self.images_download, picture_data)
        pool.map(self.download_video, video_data)
        print('[+] Done')

def main():
    parser = argparse.ArgumentParser()

    parser.add_argument('-U', '--username', help='Username or your email of your account', action='store',
    parser.add_argument('-P', '--password', help='Password of your account', action='store', required=True)
    parser.add_argument('-F', '--filename', help='Filename for storing data', action='store', required=True)
    parser.add_argument('-T', '--target', help='Profile name to search', action='store', required=True)
    args = parser.parse_args()
    with InstagramPV(args.username, args.password, Path(args.filename), args.target) as pv:

if __name__ == '__main__':

Changes :

1) Constants

2) Fixing nested dictionary in function check_availability

3) Create Static Function get_fields

Usage: myfile.py -U myemail@hotmail.com -P mypassword -F Mynamefile -T stackoverjoke

My previous comparative review tag : Scraping Instagram - Download posts, photos - videos



I don't find it necessary for these two statements to occupy four lines each:

    privacy = (

    followed_by_viewer = (

They're better off as

privacy = user.get('is_private')
followed_by_viewer = user.get('followed_by_viewer')


    media = ['graphql', 'shortcode_media', *keys]
    if list(nodes.keys())[0] == 'node':
        media = ['node', *keys]

I think would be more appropriately represented as

if list(nodes.keys())[0] == 'node':
    media = ['node', *keys]
    media = ['graphql', 'shortcode_media', *keys]


You have a class, and it's pretty reasonably laid-out now, but it probably has too many responsibilities. Consider separating it out into:

  • an InstagramScraper, containing your current
    • username
    • password
    • http_base
    • driver
    • check_availability
    • login
  • an InstagramData, containing your current
    • videos
    • pictures
    • downloading_video_images

Your InstagramScraper should not hold onto videos, pictures or even an instance of InstagramData as members. You should rework your code so that an InstagramData is constructed and returned by one method of InstagramScraper, probably calling into your other helper methods to get the necessary data.

This will make unit testing easier.

Unit Tests

It's time. Since you are serious about this project, you need tests. This is not an easy thing so will require some research and experimentation. You will want to pick a unit-testing framework - maybe nose, or maybe bare unittest, etc. Read some walkthroughs. You'll also need to rely on mocking, because you have lots of external dependencies - to requests and selenium. Once you have a few test methods in place, use a tool to measure your coverage while executing your tests. See how high you can get your coverage! You might even find some bugs during this process.

| improve this answer | |
  • \$\begingroup\$ Very nice!! Should i start, first to seperate the "god-class" and after that the unittest? The seperated classes could be inheritance? returned by one method of InstagramScraper ? With this i understood to create one class method to return the URLs of videos and images and use it for the next class something like get_all(). \$\endgroup\$ – AlexDotis Apr 19 at 16:57
  • 1
    \$\begingroup\$ Yes, do your separation before your unit testing. No, the new class should probably not inherit from your existing class. \$\endgroup\$ – Reinderien Apr 19 at 16:59
  • 1
    \$\begingroup\$ get_all() is a reasonable name. And yes, it can either return the bare URLs, or it could simply construct an InstagramData class with those URLs and return that. \$\endgroup\$ – Reinderien Apr 19 at 16:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.