Threads has a search box, but nothing to export what it finds. If you want every post that mentions your brand, a competitor or a topic in a spreadsheet, or a daily list of new mentions, you have to copy them by hand.
This tutorial shows how to scrape Threads posts by keyword with a ready-made scraper on Apify, first with clicks only, then from Python. A run that collects 80 posts takes about 20 seconds.
Disclosure: I built the scraper used here (Threads Search Scraper). It costs $2 per 1,000 posts, and Apify's free plan includes $5 of usage a month, so you can follow along without paying.
Why not the official Threads API?
Meta's Threads API does have a keyword search endpoint (/keyword_search). To search other people's posts with it you need a Meta developer app with the Threads use case, a Threads account that authorizes the app, and the threads_keyword_search permission approved through App Review. Until it is approved, the search only covers your own posts.
Once approved it's a solid option: up to 2,200 queries per 24 hours, up to 100 posts per request, and date filters back to July 2023. Its documented fields are the post's text, media type, link, timestamp and username, without like or reply counts. If you don't want to go through App Review, or you need engagement numbers, a scraper gets you there today.
What you get
For every post that matches your keyword:
- text, link, date, and author (username, display name, verified badge)
- likes, replies, reposts and quotes
- images, videos and links in the post
- language, and the post it replies to (when it is a reply)
It works without logging in to Threads: it reads the same public search the Threads website shows to logged-out visitors, so it never touches your account and only sees public posts.
Option 1: scrape Threads without code (3 minutes)
1. Open the brand-monitoring example. Go to Monitor Brand Mentions on Threads and click Try for free. Sign up for Apify if you haven't (free, no credit card). Apify copies the example into your account as a task, so you get a ready-filled form.
2. Enter your keywords. Under Search keywords the form comes prefilled with notion. Replace it with your own keyword and click Add for more (or Bulk edit to paste a list, one per line): a brand (shopify), a product, a hashtag (#booktok) or a short phrase. Each keyword is searched separately. Keep these settings:
-
Posted after:
24 hours(only posts from the last day; also accepts a date like2026-11-01) - Maximum results: a cap on the run, which is also a cap on what you pay
3. Click Save & start. When the run finishes, the Output tab shows a table of posts. Click Export to download CSV, Excel or JSON.
That run searched duolingo and shopify for the last 24 hours and returned 80 posts in 19 seconds, in English, Japanese, Chinese and many other languages.
4. Make it daily. The task from step 1 is already saved in your account. Go to Schedules in Apify Console and click Create schedule. Set the time under Schedule setup (for example every day at 9:00), add your task under Actors and tasks > Add new, and check that the schedule is Enabled. Because "Posted after" is 24 hours, each run returns the last day's posts, so a daily schedule gives you a feed of new mentions with little overlap. To get the results somewhere else automatically, connect the task to Make, Zapier, n8n or Google Drive from its Integrations tab.
Option 2: scrape Threads with Python
Install the client and pandas, and create an API token in Apify Console > Settings > API & Integrations (Create a new token; Apify shows it only once, so copy it right away):
pip install apify-client pandas
export APIFY_TOKEN=your_token
This script searches two brands for the last 24 hours, keeps the English posts and sorts them so the most discussed come first:
import os
import pandas as pd
from apify_client import ApifyClient
client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("nexuscortex/threads-search-scraper").call(run_input={
"searchQueries": ["duolingo", "shopify"],
"sort": "recent",
"dateFrom": "24 hours",
"maxPostsPerQuery": 200,
"maxItems": 400,
})
posts = pd.DataFrame(client.dataset(run.default_dataset_id).list_items().items)
# English posts only, most discussed first
english = posts[posts["language"] == "en"]
top = english.sort_values(["replyCount", "likeCount"], ascending=False)
top[["searchQuery", "createdAt", "username", "likeCount", "replyCount", "text", "url"]].to_csv("mentions.csv", index=False)
print(f"{len(posts)} posts, {len(english)} in English, saved to mentions.csv")
On the run above, 32 of the 80 posts were tagged as English. Threads leaves the language empty on many posts (32 of the 80 here), so the filter also drops some English posts; skip it if you need everything. Sorting by replies put real conversations near the top, including someone asking if anyone could build a Shopify site for them (11 replies). That's a sales lead for a Shopify freelancer, found with a plain brand keyword.
Tips for better results
Use short keywords. Threads' public search matches one or two words well and long phrases poorly. In a 7-day test, shopify developer returned 86 posts and looking for someone to build my shopify store returned none. If a phrase finds nothing, the scraper retries it word by word and keeps only posts that contain most of its words (those rows are marked matchedVia: "broadened").
Expect loose matches and filter afterwards. Threads decides what matches, and it includes posts that only mention the keyword in passing, plus replies in a matching conversation. For lead generation, search a broad keyword and filter the text in pandas for words like need, hiring or recommend, rather than searching a long phrase.
Search older periods. Set both Posted after and Posted before (2025-11-25 to 2025-11-30) to collect posts from a past campaign or event. The scraper pages back through time, so it isn't limited to the 10 to 20 posts Threads shows logged-out visitors.
Top instead of Recent. "sort": "top" with "topPeriod": "day" returns the most relevant posts of each day, which is a quick way to see what got attention about a topic this week.
Profiles too. "searchProfiles": true adds the accounts that match each keyword, with follower count and bio, which is useful for finding creators in a niche. To scrape a specific account's posts instead of a keyword, use Threads Scraper.
What it costs
You pay per post saved: $2 per 1,000 on Apify's free plan, less on paid plans. The price includes the compute the run uses, so there is no separate usage charge. Duplicates and posts outside your date range are not charged, and maxItems caps every run. Apify's free plan includes $5 of usage each month, which covers about 2,500 posts. A daily brand check that finds 50 new posts costs about $3 a month.
Limits
- Public posts only; private accounts are never accessed.
- Very busy keywords (hundreds of posts an hour) are sampled hour by hour in Recent mode, not collected in full.
- No view counts.
- Image and video links expire after a few days, so download the files you need.
- The data is public, but it is still personal data: follow Meta's terms and privacy laws such as GDPR when you store or use it.
If something breaks or you need a field that isn't there, open an issue on the Actor page and I'll take a look.


Top comments (0)