Tweet Scraper | X Scraper

Look up public X (Twitter) data: a profile, home tweets, one tweet and the replies on that page, following, and followers.

Full field reference, including HTTP examples: https://www.opentwitter.com/docs

$0.12 per 1,000 dataset rows. Apify charges one result for each row. Leave OpenTwitter API key blank. This Actor uses the key stored on it.

A name, @name, or an x.com profile link all work in Handles. Tweet ids and user ids are strings. A short number such as 1234567890 is rejected.

What each run accepts

Field	Required	Values
What to fetch	yes	tweets, profile, both, tweet, following, followers
Handles	yes, except One tweet	One account per line
Tweet id or status URL	One tweet only	15–20 digits, or a status URL
User id	no	Numeric id you already know. Skips one profile lookup
Count	see each section	Home tweets: 20–40, default 20. Following and followers: 1–200, default 200
Concurrency	no	How many accounts at once. Default 3, maximum 8
API base URL	no	Leave https://api.opentwitter.com

Each successful item is one dataset row. Open Storage to see profile, tweet, or user. The table shows handle, endpoint, status, and ok.

A missing or suspended account is HTTP 404 and still one row:

{
  "handle": "elonmusk",
  "endpoint": "profile",
  "status": 404,
  "ok": false,
  "error": "This account was not found"
}

A bad key, a bad parameter, no balance, or a server error is not written to the dataset and is not charged.

Profile

One public profile. Count is ignored. One profile is 1 result. A 404 is also 1 result.

Request:

{ "endpoint": "profile", "handles": ["elonmusk"] }

Calls POST https://api.opentwitter.com/v1/lookup/profile with { "handle": "elonmusk" }.

Response, one row:

{
  "handle": "elonmusk",
  "endpoint": "profile",
  "status": 200,
  "ok": true,
  "profile": {
    "user_id": "44196397",
    "screen_name": "elonmusk",
    "display_name": "Elon Musk",
    "bio": "",
    "location": "",
    "website": "",
    "followers_count": 0,
    "following_count": 0,
    "tweets_count": 0,
    "likes_count": 0,
    "verified": false,
    "protected": false,
    "avatar_url": "https://pbs.twimg.com/profile_images/1.jpg",
    "banner_url": "https://pbs.twimg.com/profile_banners/1.jpg",
    "created_at": "Tue Jun 02 00:00:00 +0000 2009"
  }
}

Home tweets

Latest posts on the profile timeline. count must be 20–40. Default 20. One request to X returns at least 20 posts, so a smaller count is rejected. Each returned tweet is 1 result. Duplicate tweet ids are removed. Zero tweets, or a 404, is still 1 result.

Request:

{ "endpoint": "tweets", "handles": ["elonmusk"], "count": 20 }

Calls POST https://api.opentwitter.com/v1/lookup/tweets with { "handle": "elonmusk", "count": 20 }. Add "user_id": "44196397" when you already know it.

Response, one row per tweet:

{
  "handle": "elonmusk",
  "endpoint": "tweets",
  "status": 200,
  "ok": true,
  "tweet": {
    "tweet_id": "1840000000000000002",
    "text": "tweet text",
    "created_at": "2026-09-17T10:00:00.000Z",
    "url": "https://x.com/elonmusk/status/1840000000000000002",
    "likes": 0,
    "retweets": 0,
    "replies": 0,
    "views": 0,
    "is_retweet": false,
    "is_reply": false,
    "is_quoted": false,
    "urls": []
  }
}

Several accounts. Default concurrency is 3, maximum 8.

{ "endpoint": "tweets", "handles": ["elonmusk", "nasa"], "count": 20, "concurrency": 3 }

Profile and tweets

Both in one run. Tweet count is 20–40, same floor as home tweets. The dataset is split: one profile row, then one row per tweet. Each side is billed on its own. One side can fail while the other succeeds. A protected account can still return the profile, with tweets as a 404 row.

Request:

{ "endpoint": "both", "handles": ["elonmusk"], "count": 20 }

Calls POST https://api.opentwitter.com/v1/lookup with { "handle": "elonmusk", "count": 20 }.

Response rows use the same profile object as Profile and the same tweet object as Home tweets. endpoint on those rows is profile or tweets.

One tweet

That tweet, plus the replies on the same page. Handles and count are ignored. The main tweet and each reply are each 1 result. Duplicate ids are removed. Zero items, or a 404, is still 1 result.

Request. Keep the id a string. A status URL is accepted.

{ "endpoint": "tweet", "tweetId": "1840000000000000002" }

{ "endpoint": "tweet", "tweetId": "https://x.com/elonmusk/status/1840000000000000002" }

Calls POST https://api.opentwitter.com/v1/lookup/tweet with { "tweet_id": "1840000000000000002" }.

Response, one row for the tweet and one row for each reply. The shape is the same tweet object as Home tweets. A reply has "is_reply": true.

Following

Accounts this user follows, most recent first. count is 1–200, default 200. The lookup stops at that count, so the bill matches the accounts returned. Each account is 1 result. Duplicate user ids are removed. Zero accounts, or a 404, is still 1 result. X does not return the time they were followed. account_created_at is when that account was created.

Request:

{ "endpoint": "following", "handles": ["elonmusk"], "count": 50 }

Calls POST https://api.opentwitter.com/v1/lookup/following with { "handle": "elonmusk", "count": 50 }.

Response, one row per account:

{
  "handle": "elonmusk",
  "endpoint": "following",
  "status": 200,
  "ok": true,
  "user": {
    "user_id": "12",
    "handle": "jack",
    "name": "jack",
    "bio": "",
    "followers_count": 0,
    "following_count": 0,
    "tweets_count": 0,
    "verified": false,
    "protected": false,
    "avatar_url": null,
    "location": "",
    "account_created_at": "Tue Mar 21 00:00:00 +0000 2006"
  }
}

This run does not page further than count. next_cursor from the API is not returned as its own field.

Followers

Accounts that follow this user. Same count rule and same user object as Following.

Request:

{ "endpoint": "followers", "handles": ["elonmusk"], "count": 50 }

Calls POST https://api.opentwitter.com/v1/lookup/followers with { "handle": "elonmusk", "count": 50 }.

Response rows use "endpoint": "followers" and the same user object as Following.

Not included

No keyword search, no lists, no date range, and no batch endpoint. More than 40 home tweets, or more than 200 follow-list accounts, needs another run. This Actor does not monitor an account over time.