关于Instagram接口更新及媒体评论端点变更的技术咨询

Hey there, I’ve been dealing with exactly this Instagram scraping issue lately, so let me break down what I’ve figured out:

• Non-negotiable rule now: Ever since the Cambridge Analytica incident fallout, Instagram has locked down all its endpoints—every single request has to be authenticated with a logged-in user session. No more anonymous scraping; you need to attach valid session cookies to your request headers.
• Comment endpoint specific changes:
  • Old (now invalid) media comment endpoint:

    https://www.instagram.com/graphql/query/?query_id=17888483320059182&id={0}&first=100&after={1}
    

  • New endpoint details: Instagram revised the GraphQL endpoint structure, and the key addition is that you must include your logged-in cookie (containing sessionid and other critical auth tokens) in the request headers. Also, the query_id value might have changed—your best bet is to capture a live request from a logged-in browser to get the current valid query_id and full endpoint path.

Quick Practical Tips #

• To get a valid session: Use tools like Selenium or Playwright to simulate a manual Instagram login, then save the resulting cookies. Attach these cookies to every subsequent scraping request.
• Anti-scraping heads-up: Instagram’s bot detection is tough—pace your requests like a real user, avoid bulk scraping, and regularly refresh your session if you start getting blocked. Also, query_id can get updated without warning, so double-check it periodically by re-capturing a browser request.

内容的提问来源于stack exchange，提问作者supernova