Threads Search & Replies Scraper

Find out what people are saying about your brand on Threads, and read whole conversations under a post.

Give it a list of keywords or hashtags and you get back the public Threads posts that match, each with the author, the text, the timestamp and the like, reply, repost and quote counts. Paste a post link instead and you get that post plus the replies Threads shows publicly underneath it, one row per reply.

There is no account to connect. The scraper reads the same public, logged-out view of Threads that anyone gets in a browser window without signing in, so there are no cookies to refresh, no session to get banned, and no API key to buy.

What you get

One flat row per post or reply.

Field	What it is
recordType	post for a top-level post, reply for a reply under one
postId, shortcode, url	Identifiers and the permalink
text	The full post text, emoji and line breaks kept
postedAt	ISO 8601 UTC timestamp
language	Language Threads detected, when it reports one
authorUsername, authorName, authorId	Who posted it
authorIsVerified, authorIsPrivate	Account flags
authorUrl, authorAvatarUrl	Profile link and picture
likeCount, replyCount, repostCount, quoteCount, reshareCount	Engagement
isReply, replyToUsername, rootPostUrl	Where the row sits in a thread
mediaType, mediaUrls	text, image, video or carousel, plus the file URLs
mentions, hashtags, links	Accounts tagged, tags used, outbound URLs
isPaidPartnership	Set when Threads labels the post as sponsored
query, sourceUrl, scrapedAt	Which input produced the row, and when

Empty means empty. When a field is genuinely absent (an author who hid their like counts, a post with no media) you get null, never a zero standing in for "we could not tell".

Who it is for

Social listening and brand monitoring. Put your brand, your product names and your competitors in the keyword list, run it on a schedule, and you have a feed of every public Threads post that mentions any of them, ranked by the engagement they actually got. The query field tells you which term matched, so one run can cover a whole watchlist.

Community and support teams. Paste the link to a post that is blowing up and get every public reply as structured rows, with the author and the like count on each one. Much faster than scrolling a thread and copying quotes by hand.

Market and content research. Pull the posts around a topic or a hashtag, look at what the high-engagement ones have in common, and export it to a spreadsheet or a notebook. Hashtags and mentions come pre-extracted.

Influencer and creator sourcing. Search your niche, sort by likes, and you have a shortlist of accounts already posting about it, with their verified status and follower-facing profile links.

Input examples

Track a brand and its competitors:

{
  "keywords": ["nike running", "adidas running", "#runningshoes"],
  "maxResults": 300
}

Pull the replies under two specific posts:

{
  "postUrls": [
    "https://www.threads.com/@claudeai/post/DVPbBO5icc6",
    "https://www.threads.com/@zuck/post/DMZ3zK8Rvzg"
  ],
  "maxResults": 200
}

Both at once, which is the usual monitoring setup:

{
  "keywords": ["openai", "anthropic"],
  "postUrls": ["https://www.threads.com/@claudeai/post/DVPbBO5icc6"],
  "maxResults": 500
}

You can also paste a Threads search link (https://www.threads.com/search?q=...) or a tag page (https://www.threads.com/t/ai) straight into the keyword list.

Pricing

You pay per post or reply delivered. Nothing else.

Event	Price
Post or reply	$0.002 per row ($2.00 per 1,000)
Actor start	$0.001 per run

Bronze, Silver and Gold plans get 10%, 20% and 30% off the per-row price, so 1,000 posts are $2.00, or $1.40 on Gold. The start fee is flat.

Apify platform usage is included in the price — and this actor uses no proxy at all, so there is nothing else to pay: no compute line, no proxy line, no subscription, no API key. A keyword that matches nothing costs the start fee and no more.

Frequently asked questions

Do I need a Threads or Instagram login? No. No login, no cookies, no session token, no API key. The scraper only reads what Threads serves to a signed-out visitor.

How many posts do I get per keyword? Threads gives a signed-out reader one batch of results per query rather than endless scroll, so expect roughly 20 to 30 posts per keyword. The scraper already squeezes that as far as it goes by searching both the posts tab and the tag tab for every keyword and merging the two, which is measurably more than either tab alone. If you want more volume on a topic, give it more keyword variants: six phrasings around "nike" returned 88 distinct posts in testing, where one phrasing returned 27.

How many replies do I get per post? The replies Threads renders publicly under the post, which in testing was around 27 on a thread with several hundred total replies. Nobody gets the full reply list of a big thread without an account, and any scraper that promises one is either logging in on your behalf or overselling. This one tells you what it actually returns.

Why do I get fewer results than maxResults? maxResults is a ceiling, not a target. If the public data is thinner than your ceiling, you get what exists and pay only for the rows you received.

Are replies included in keyword search results? Yes, and they are labelled. A reply that mentions your brand is still a brand mention, so it comes back with recordType set to reply and you can filter it out if you would rather only see top-level posts.

Can I schedule it? Yes. Set up an Apify schedule and point it at the same keyword list. Rows carry postId and shortcode, so de-duplicating across runs on your side is a single join.

Does it work on private accounts? No, and it should not. Private posts are not part of the public view and this scraper never tries to reach them.

What happens if a post link is dead? The run logs it, skips it, and carries on with the rest of your input. A bad link never fails the whole run.

Can I get profile details and follower counts? That is a different job and there is a very good actor for it already. Use Meta Threads Profile Scraper for profile info, and this one for posts, keywords and replies.

Notes on reliability

The scraper impersonates a real browser at the TLS level rather than sending browser-shaped headers over a Node handshake, which is what Meta actually checks. Three browser profiles are tried in rotation, and every attempt is logged with its status code so a failure is diagnosable from the run log instead of showing up as a mysteriously empty dataset. A run that finds nothing says so in its status message rather than quietly succeeding with an empty result.

Every Apify actor publishes its own 30-day run stats, and on the figures read on 29 August 2026 the largest Threads scraper on the Store had failed 13.7% of its 63,993 runs, while this one had failed none of its 197. That is a much smaller sample and it will only be a fair comparison once this actor has run as much — read both records yourself before trusting either, because they move.

Related actors

Brand mentions do not live on one network, so two siblings cover the rest:

• Telegram Channel Monitor — the same keyword watch across public Telegram channels, also with no login and no API key, priced per message delivered.
• Brand News Monitor — the press side: news articles mentioning a brand across several sources at once, for when a Threads spike needs a cause.

---

Keywords: threads scraper, meta threads scraper, threads keyword search, threads hashtag scraper, threads replies scraper, threads comments scraper, threads brand monitoring, threads social listening, threads post scraper, scrape threads without login, threads.net scraper, threads mentions tracker.