Instagram Scraper

Fetch structured Instagram data through a single Actor. Pick an action and provide its inputs — the Actor stores the result in the dataset, one item per post, comment, user, story or highlight.

What you can fetch

Action	Fetches	Required input	Optional
userInfo	User profile / account info	userId or username	—
stories	Active stories	userId or username	—
highlights	Highlight tray	userId or username	—
highlightStories	Items inside a highlight	highlightId	—
posts	Timeline posts	userId or username	maxId
reels	Reels	userId or username	maxId
taggedPosts	Tagged posts	userId or username	maxId
followers	Followers	userId or username	maxId
followings	Followings	userId or username	maxId, count
likers	Users who liked a post	url or shortcode	—
comments	Comments on a post	url or shortcode	maxId

Input

Provide action plus the fields it needs (see the table above). Example:

{ "action": "posts", "username": "nasa" }

{ "action": "comments", "shortcode": "C1a2b3c4d5", "maxId": "..." }

Output

Each run writes one page of results:

Response	Dataset	Key-value store
List actions (all except userInfo)	one item per list element (post, reel, comment, user, story, highlight); each item has a copy of the cursor in _pageInfo	PAGE_INFO record with the cursor
Empty list	no items	PAGE_INFO record is still written
userInfo	one item: the profile	—
Error	one item { "success": false, "message": ... }	—

Example item of posts:

{
  "id": "4001971394119634476",
  "type": "image",
  "shortcode": "DeJ28eTxo4s",
  "permalink": "https://www.instagram.com/p/DeJ28eTxo4s/",
  "caption": "...",
  "takenAt": 1791292211,
  "likeCount": 120000,
  "commentCount": 900,
  "media": [ { "type": "image", "url": "https://...", "thumbnailUrl": "https://..." } ],
  "owner": { "userId": "528817151", "username": "nasa" },
  "_pageInfo": { "hasNextPage": true, "maxId": "4002789380695323831_528817151" }
}

• Fields are camelCase. A field that has no value is omitted (no empty strings).
• Times (takenAt, createdAt) are Unix timestamps in seconds.
• Other top-level fields of a list response are copied into every item with a _ prefix: stories items carry the account in _owner.
• In Apify Console the dataset has a view (table) for every action and an Errors view (.actor/dataset_schema.json); pick the view of the action you ran. The Output tab links to the dataset items and to the PAGE_INFO record (.actor/output_schema.json).

Pagination

Paginated actions (posts, reels, taggedPosts, followers, followings, comments) return a cursor { "hasNextPage": true, "maxId": "..." }. It is stored twice:

• in the PAGE_INFO record of the run's default key-value store;
• in the _pageInfo field of every dataset item of that page.

To fetch the next page, run the Actor again and pass maxId as input maxId. There is no maxId when hasNextPage is false. Each run returns a single page. If a page is empty there are no items, so read the cursor from PAGE_INFO.

Run via API

curl -X POST \
  "https://api.apify.com/v2/acts/<ORG>~instagram/run-sync-get-dataset-items?token=<APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{ "action": "posts", "username": "nasa" }'

The response is an array of dataset items — one per post. The cursor for the next page is in items[0]._pageInfo.maxId, so one request is enough per page:

curl -X POST \
  "https://api.apify.com/v2/acts/<ORG>~instagram/run-sync-get-dataset-items?token=<APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{ "action": "posts", "username": "nasa", "maxId": "<items[0]._pageInfo.maxId>" }'

For userInfo the array has one item, the profile. An error is one item with "success": false.

Pricing

This Actor uses pay-per-event pricing — you pay only for these events, platform usage (compute) is included:

Event	Price
Result — one dataset item	$2.50 per 1,000 results ($0.0025 each)
Actor start — per GB of run memory, at least one	$0.001

An error is returned as one dataset item, so it counts as one result.

Notes

• Validation happens before any network call: an unknown action or a missing required field returns a clear error.
• Business errors (e.g. "page not found", private account) are returned as one dataset item as-is ({ "success": false, "message": ... }) and the run still succeeds, so you can handle them programmatically.