Threads Account Replies Scraper

Extract every reply a Threads user has posted to other people's threads — the full Replies tab, complete with the parent post each reply is answering — in a clean structured JSON output.

Why use this actor

• No account / no login required — just give it a Threads username or profile URL.
• No API key needed — Threads has no public consumer API; this actor returns the same data the web app shows on the Replies tab.
• Reply chain context — every record bundles the user's reply together with the original parent thread it answered, so you can see the conversation, not just the response in isolation.
• Engagement on replies — like counts, codes, and timestamps are returned per reply, so you can rank an account's most-engaged replies (a very different signal from their own posts).
• Bulk input — pass a list of usernames or URLs in one run; results stream straight into your dataset.
• Stable JSON output suitable for pipelines, spreadsheets, and databases — every row carries _input, _source, _scrapedAt envelope fields so you can join results back to your input list.

How it works

1. You provide a list of Threads usernames (e.g. mosseri) or full profile URLs.
2. The actor opens each profile, reads the public Replies tab, and paginates through it until it hits your maxPosts cap (or the end of the list).
3. For every reply it returns the reply post itself plus the parent thread it was answering, packaged in a flat JSON record.
4. Results stream into your dataset, ready to download as JSON, CSV, or Excel.

You do not need to manage scrapers, browsers, or rotating IPs — all handled internally.

Input

{
  "usernames": [
    "mosseri",
    "zuck"
  ],
  "maxPosts": 30,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["DATACENTER"]
  }
}

Field	Type	Description
usernames	array	List of Threads usernames or full profile URLs to scrape. Both mosseri and https://www.threads.com/@mosseri are accepted.
maxPosts	integer	Maximum replies returned per username. Default 30. Set to 0 for unlimited.
proxyConfiguration	object	Apify Proxy settings. Datacenter proxy works for most cases; use Residential for very large batches.

Output

Input: mosseri, maxPosts: 3

Each result is one reply chain: the parent thread the user was responding to, followed by the user's own reply. The example below shows two chains — Adam Mosseri (mosseri) replying to luizgcribeiro about the new Threads logo, then replying to theonion.

[
  {
    "_input": "mosseri",
    "_source": "S1-primary",
    "_scrapedAt": "2026-05-18T10:59:33.621591+00:00",
    "id": "3897787543714767562",
    "thread_items": [
      {
        "post": {
          "pk": "3897691905908415945",
          "id": "3897691905908415945_63453918378",
          "code": "DYXYiM3lh3J",
          "taken_at": 1778861104,
          "like_count": 504,
          "caption": { "text": "new Threads logo swag just dropped" },
          "user": {
            "pk": "63453918378",
            "username": "luizgcribeiro",
            "is_verified": false,
            "profile_pic_url": "https://instagram.fcgk12-2.fna.fbcdn.net/v/t51.82787-19/632836115_..."
          },
          "image_versions2": { "candidates": [ { "height": 3024, "width": 4032, "url": "https://..." } ] },
          "text_post_app_info": {
            "reply_to_author": null,
            "is_post_unavailable": false,
            "share_info": { "quoted_post": null, "reposted_post": null }
          }
        },
        "line_type": "none",
        "should_show_replies_cta": true,
        "__typename": "XDTThreadItem"
      },
      {
        "post": {
          "pk": "3897787543714767562",
          "id": "3897787543714767562_63482099442",
          "code": "DYXuR6hD6LK",
          "taken_at": 1778872505,
          "like_count": 64,
          "caption": { "text": "👌🏼" },
          "user": {
            "pk": "63482099442",
            "username": "mosseri",
            "is_verified": true,
            "profile_pic_url": "https://instagram.fcgk12-1.fna.fbcdn.net/v/t51.82787-19/652081753_..."
          },
          "text_post_app_info": {
            "reply_to_author": { "username": "luizgcribeiro", "id": "63453918378" },
            "is_post_unavailable": false,
            "share_info": { "quoted_post": null, "reposted_post": null }
          }
        },
        "line_type": "none",
        "should_show_replies_cta": true,
        "__typename": "XDTThreadItem"
      }
    ]
  },
  {
    "_input": "mosseri",
    "_source": "S1-primary",
    "_scrapedAt": "2026-05-18T10:59:33.622601+00:00",
    "id": "3895875287850949047",
    "thread_items": [
      {
        "post": {
          "pk": "3895681172517134037",
          "code": "DYQPWLeD57V",
          "taken_at": 1778621410,
          "like_count": 203,
          "caption": { "text": "Celtics Attempt To Lure Giannis By Announcing Payton Pritchard His Brother Now https://theonion.com/..." },
          "user": { "pk": "63484314604", "username": "theonion", "is_verified": true },
          "text_post_app_info": { "reply_to_author": null, "is_post_unavailable": false }
        },
        "__typename": "XDTThreadItem"
      },
      {
        "post": {
          "pk": "3895875287850949047",
          "code": "DYQ7e7dlBm3",
          "taken_at": 1778644546,
          "like_count": 18,
          "caption": { "text": "love this" },
          "user": { "pk": "63482099442", "username": "mosseri", "is_verified": true },
          "text_post_app_info": {
            "reply_to_author": { "username": "theonion", "id": "63484314604" },
            "is_post_unavailable": false
          }
        },
        "__typename": "XDTThreadItem"
      }
    ]
  }
]

Field	Type	Description
_input	string	The username or URL exactly as you supplied it. Use this to join results back to your input list.
_source	string	Internal tag for the path used to fetch the record. S1-primary means the fastest, richest path; values starting with S2- indicate a fallback was used.
_scrapedAt	string	ISO-8601 UTC timestamp when the record was scraped.
id	string	ID of the user's own reply (the last item in thread_items).
thread_items	array	The full reply chain in chronological order. The last item is the user's reply; earlier items are the parent thread(s) being answered.
thread_items[].post.pk	string	Numeric post ID for that item in the chain.
thread_items[].post.code	string	Short URL slug; the full post URL is https://www.threads.com/@<user>/post/<code>.
thread_items[].post.taken_at	integer	Unix timestamp (seconds) when the post was published.
thread_items[].post.like_count	integer	Number of likes on that specific item.
thread_items[].post.caption.text	string	The post's text body. null if it is an image-only post.
thread_items[].post.user.username	string	Author handle of that item. Match this to your input username to identify which row is the user's own reply.
thread_items[].post.user.pk	string	Numeric user ID of the post's author. Stable across handle changes.
thread_items[].post.user.is_verified	boolean	true if the author has a verified badge.
thread_items[].post.image_versions2.candidates	array	Image renditions attached to the post (multiple sizes from 150px up to original). Empty when the post is text-only.
thread_items[].post.video_versions	array	Video renditions; null for non-video posts.
thread_items[].post.carousel_media	array	Carousel slides; null when the post is single-media or text-only.
thread_items[].post.text_post_app_info.reply_to_author	object	If this item is itself a reply, the parent author's username + numeric id. null on top-level posts. Use this field to walk the conversation chain.
thread_items[].post.text_post_app_info.share_info.quoted_post	object	Populated when the item quotes another thread; otherwise null.
thread_items[].post.text_post_app_info.share_info.reposted_post	object	Populated when the item is a repost; otherwise null.
thread_items[].post.text_post_app_info.is_post_unavailable	boolean	true if the parent thread was deleted or hidden but the reply still references it.

Error envelope

Profiles that can't be resolved or replies that fail to load return a structured error instead of crashing the run:

{
  "_input": "this-user-does-not-exist-xyz",
  "_source": "S2-fallback",
  "_error": "user_resolve_failed",
  "_scrapedAt": "2026-05-18T10:59:34.000000+00:00"
}

Filter on _error to triage failed rows. Common values: user_resolve_failed (the username does not exist or is private) and fetch_failed (the Replies tab could not be loaded — usually transient).

Pricing

This actor is billed per result: $6.00 per 1,000 reply chains (Tier 4). Each reply chain returned = 1 result. Error rows (user_resolve_failed, fetch_failed) are not billed.

Other Sosmed Actors

Platform	Actor	Best for
Threads	Threads Account Scraper	Profile data (bio, followers, verified) for a Threads handle
Threads	Threads Account Threads Scraper	The user's own posts (Threads tab, not replies)
Threads	Threads Post Detail Scraper	Full conversation tree under any single thread URL
YouTube	YouTube Video Comments Scraper	Comment threads on any public YouTube video
Reddit	Reddit User Profile Scraper	Karma, posts, trophies for any Reddit account
X / Twitter	X Account Tweets Scraper	Latest tweets from any public X handle
Bluesky	Bluesky Post Detail Scraper	Replies and quotes under any Bluesky post

Browse the full catalog at apify.com/xtracto.

Notes

• Only public replies are returned. If the account is private, the run yields a single _error: "user_resolve_failed" row for that username.
• Parent threads inside thread_items may be deleted or hidden by the time of scraping; in that case the parent's text_post_app_info.is_post_unavailable is true and the parent author's content fields may be empty. The user's own reply is still returned.
• Conversation chain depth varies. Most reply chains are 2 items long (parent + reply), but threads with nested context can return 3+ items in thread_items — always treat the last item as the user's own reply and earlier items as ancestors.
• The numeric pk and the user.pk are the most reliable identifiers — Threads handles can be changed, but these numeric IDs do not.
• For large jobs (>200 usernames), switch proxyConfiguration.apifyProxyGroups to ["RESIDENTIAL"] to avoid per-IP rate limits.