Instagram GraphQL API

Instagram's website doesn't use a public REST API. When you open a profile or scroll a feed, the browser sends POST requests to instagram.com/graphql/query (older builds used GET /graphql/query with a query_hash). Instead of sending a GraphQL query string, the client sends the ID of a persisted query plus a JSON blob of variables.

POST https://www.instagram.com/graphql/query
Content-Type: application/x-www-form-urlencoded
X-IG-App-ID: <web app id>
X-CSRFToken: <csrftoken cookie>
Cookie: csrftoken=...; sessionid=...; ...

doc_id=<numeric persisted query id>
&variables={"username":"nike", ...}

The response is a deeply nested GraphQL payload: data.user.edge_owner_to_timeline_media.edges[].node and similar, with pagination in page_info.end_cursor.

Why GraphQL scrapers keep breaking

• ▹doc_id and query_hash values are build artifacts. They change when Instagram ships a new web bundle, and old IDs start returning errors.
• ▹Most queries now need a logged-in session. Anonymous requests get redirected to the login wall or rate-limited after a handful of calls.
• ▹Logged-in scraping puts the account at risk: checkpoints, challenges, and bans.
• ▹Rate limits are per IP and per session, so scale means proxy pools and account rotation.
• ▹Response shapes drift: fields move between edges, get renamed, or disappear without notice.

The same data over a stable REST API

This API sits in front of all of that. You call a versioned endpoint with an API key, and it handles sessions, proxies, query IDs, and retries, then returns flat JSON. Here's a profile's recent posts, the equivalent of the timeline-media GraphQL query:

import os, requests

r = requests.get(
    "https://api.profilequery.com/v1/profile/posts",
    params={"handle": "nike"},
    headers={"Authorization": f"Bearer {os.environ['INSTAGRAMAPI_KEY']}"},
)
data = r.json()["data"]
for post in data["items"]:
    print(post["shortcode"], post["like_count"], post["comment_count"])

# next page: pass data["next_cursor"] back as cursor=

GraphQL query → REST endpoint

• ▹Profile info (user by username) → GET /v1/profile?handle=…
• ▹Timeline media (profile posts) → GET /v1/profile/posts?handle=…&cursor=…
• ▹Clips / reels tab → GET /v1/profile/reels?handle=…
• ▹Followers / following edges → GET /v1/profile/followers, /v1/profile/following
• ▹Post / shortcode media → GET /v1/post?post=…
• ▹Comments on a post → GET /v1/post/comments?post=…
• ▹Likers → GET /v1/post/likers?post=…
• ▹Hashtag top / recent media → GET /v1/hashtag/top, /v1/hashtag/recent