Library
Scrapers
29 prebuilt collectors, sorted by what they collect from. Fill in a form, pick the proxy the requests leave through, and the rows land in a table of their own — with no browser window anywhere.
On this page
What a scraper is
A scraper has one job and a very small contract: input in, one table out, and no browser at all. You point it at something public — a search, an account, a place — and the rows are the whole deliverable.
That is what separates it from an automation, which drives a real browser window on one of your profiles and can do anything. A workflow that operates your own accounts is an automation. A collector pointed at somebody else’s public page is a scraper, and it runs over plain HTTP with a real browser’s TLS fingerprint — which is why a Google Maps search that used to take eight minutes in a window now takes about six seconds in none.
Scrapers cost none of your plan’s automation slots, they never open a browser window, and all but 2 of them collect signed out. The exceptions are listed below, they are yours to connect or not, and every row a scraper writes records which way it was collected.
The 29 scrapers
One card per job rather than per service — a service usually has several, because its profiles, its posts and the people who follow one are three different collections with three different forms. Everything below is read from the app’s own catalogue, so this list is what ships rather than a description of it. Each name goes to that card’s own page — its real form, and every column the table it writes is created with.
| Scraper | You give it | Columns | Account | Proxy |
|---|---|---|---|---|
| Google Keyword Ideas | Seeds | 7 | No | Optional |
| YouTube Keyword Ideas | Seeds | 7 | No | Optional |
| Bing Keyword Ideas | Seeds | 6 | No | Optional |
| Amazon Keyword Ideas | Seeds | 6 | No | Optional |
| Google Trending Now | Countries | 11 | No | Optional |
| YouTube Search Results | Keywords | 14 | No | Optional |
| Google Search Results | Keywords | 9 | No | Optional |
| Google Trends | Keywords | 7 | No | Optional |
| YouTube Transcripts | Videos | 10 | No | Optional |
| Google Maps | What to find | 47 | No | Optional |
| Instagram Profiles | Accounts | 34 | Yours, once | Advised |
| Instagram Posts | Accounts | 27 | No | Advised |
| Instagram Followers | Accounts | 13 | Yours, once | Advised |
| LinkedIn Profiles | Profiles | 21 | No | Optional |
| LinkedIn Companies | Companies | 29 | No | Optional |
| LinkedIn Company Posts | Companies | 11 | No | Optional |
| LinkedIn Posts | Posts | 21 | No | Optional |
| LinkedIn Post Comments | Posts | 10 | No | Optional |
| TikTok Profiles | Accounts | 24 | No | Advised |
| TikTok Profile Videos | Accounts | 15 | No | Advised |
| TikTok Videos | Video URLs | 39 | No | Advised |
| X Profiles | Accounts | 24 | No | Optional |
| X Posts | Accounts | 27 | No | Advised |
| X Post Lookup | Posts | 19 | No | Optional |
| Shopify Storefronts | Stores | 22 | No | Optional |
| Indeed Jobs | Searches | 27 | No | Advised |
| Greenhouse Job Boards | Boards | 16 | No | Optional |
| Ashby Job Boards | Boards | 21 | No | Optional |
| Lever Job Boards | Boards | 15 | No | Optional |
Search & SEO
What people are typing into search, what is trending today, and what already ranks for it.
Every phrase Google offers to finish your seed with — expanded across the alphabet, the question words and the buying-intent words, ranked in Google’s own order. A few hundred real searches from one word, with no account and no browser.
- You give it
- Seeds
- You get back
- 7 columns, from 10 settings
- Browser
- Not needed
What YouTube offers to finish your seed with — a different list from Google’s for the same words, because people arrive at YouTube to be shown how rather than to buy. The titles and topics a channel should be making, ranked in YouTube’s own order.
- You give it
- Seeds
- You get back
- 7 columns, from 10 settings
- Browser
- Not needed
The same expansion run against Bing’s autocomplete — which is also DuckDuckGo’s. A cheap second opinion on a seed: a phrase both engines suggest is a phrase people really type, and one only Bing has is often a gap worth taking.
- You give it
- Seeds
- You get back
- 6 columns, from 9 settings
- Browser
- Not needed
What shoppers type into Amazon around your seed. The highest buying intent of any keyword source — every one of these searches was somebody about to spend money — and the phrasing is product-shaped rather than question-shaped.
- You give it
- Seeds
- You get back
- 6 columns, from 9 settings
- Browser
- Not needed
What a country is searching for today that it was not searching for yesterday — with Google’s own rough volume and the news story behind each one. Run it on a schedule and the table becomes a record of what broke, and when.
- You give it
- Countries
- You get back
- 11 columns, from 1 setting
- Browser
- Not needed
The videos that rank for a keyword, with views, channel, length and age. Feed it the keywords the Ideas cards found and the table answers the question those cards cannot: is anyone already serving this, and how well.
- You give it
- Keywords
- You get back
- 14 columns, from 6 settings
- Browser
- Not needed
The pages ranking for a keyword, in order, with their titles and snippets. What every rank tracker is underneath — point it at the keywords the Ideas cards found and you have the competition for each one.
- You give it
- Keywords
- You get back
- 9 columns, from 4 settings
- Browser
- Not needed
The queries rising around a keyword, with Trends’ own interest score. The one card here that reports measured demand rather than suggestion order — use it to tell a keyword that is growing from one that is merely popular.
- You give it
- Keywords
- You get back
- 7 columns, from 4 settings
- Browser
- Not needed
Every line spoken in a video, timed, one row per caption. Point it at the videos already ranking for your keyword and the table is what they cover — the raw material for a script that covers it better.
- You give it
- Videos
- You get back
- 10 columns, from 2 settings
- Browser
- Not needed
Places
Businesses and locations off the maps and directories that list them.
Every place a Maps search returns, filed as rows with no browser window: name, category, full postal address, phone, website, rating, reviews and — when you ask for it — the email address off the business’s own site.
- You give it
- What to find
- You get back
- 47 columns, from 13 settings
- Browser
- Not needed
Social
Public profiles, the posts they have published, and who follows them.
The full header behind any public account — bio, follower and post counts, verification, business category and the link in the bio — and, when you ask for it, the email address off the site that link points at.
- You give it
- Accounts
- You get back
- 34 columns, from 2 settings
- Browser
- Once, to sign in
Instagram Posts
Proxy advisedEverything a public account has posted — posts, carousels and reels — with captions, timestamps, likes, comments, views and the media links, filed newest first with a row per post.
- You give it
- Accounts
- You get back
- 27 columns, from 4 settings
- Browser
- Not needed
Who follows an account, or who it follows back — handle, name, verification and Instagram’s own id for each one, a row per account and ready to join onto a profiles table.
- You give it
- Accounts
- You get back
- 13 columns, from 3 settings
- Browser
- Once, to sign in
The member behind any public profile URL — name, summary, city, current employer, schools and follower count, one row per person, read from the page LinkedIn serves without an account.
- You give it
- Profiles
- You get back
- 21 columns, from 5 settings
- Browser
- Not needed
The organisation behind any public LinkedIn company page — name, description, tagline, live employee count, website and registered city, one row per company, read from the page LinkedIn serves without an account.
- You give it
- Companies
- You get back
- 29 columns, from 6 settings
- Browser
- Not needed
The recent posts on any public LinkedIn company page — full body text, headline, publication time and the post’s own URL, one row per post. LinkedIn embeds seven to ten of them for a signed-out reader and walls the rest, so that is what this collects.
- You give it
- Companies
- You get back
- 11 columns, from 4 settings
- Browser
- Not needed
The post behind any public LinkedIn post URL — full body, author, time, total reactions, comment count and any attached media, one row per post, read from the page LinkedIn serves without an account.
- You give it
- Posts
- You get back
- 21 columns, from 6 settings
- Browser
- Not needed
The people who publicly commented on any LinkedIn post — name, what they wrote, when, and how many likes it drew, one row per comment. LinkedIn shows a signed-out reader the first eight to ten and walls the rest, and it never shows who merely reacted.
- You give it
- Posts
- You get back
- 10 columns, from 6 settings
- Browser
- Not needed
TikTok Profiles
Proxy advisedThe account behind any public TikTok @handle — name, bio, link in bio, follower, like and video counts, one row per account, read from the page TikTok serves without an account.
- You give it
- Accounts
- You get back
- 24 columns, from 6 settings
- Browser
- Not needed
TikTok Profile Videos
Proxy advisedThe ten most recent videos on any public TikTok account — caption, plays, cover and the video’s own address, one row per video. Ten is TikTok’s limit for a signed-out reader, not this card’s.
- You give it
- Accounts
- You get back
- 15 columns, from 4 settings
- Browser
- Not needed
TikTok Videos
Proxy advisedEverything TikTok publishes about a video you have the address of — plays, likes, comments, shares, saves and reposts, the caption and its hashtags, the sound, the author and their counts, one row per video.
- You give it
- Video URLs
- You get back
- 39 columns, from 6 settings
- Browser
- Not needed
The account behind a handle — bio, follower and following counts, post count, website, location, join date and X’s own id, a row per account and ready to join onto a posts table.
- You give it
- Accounts
- You get back
- 24 columns, from 1 setting
- Browser
- Not needed
X Posts
Proxy advisedEverything an account has posted — text, timestamps, likes, reposts, replies, quotes, bookmarks, hashtags, mentions and the links behind the t.co wrappers, a row per post.
- You give it
- Accounts
- You get back
- 27 columns, from 6 settings
- Browser
- Not needed
Read named posts by link or id — text, author, likes and how big the thread under them got. One request each, and the only card in the catalogue that needs no credential of any kind.
- You give it
- Posts
- You get back
- 19 columns, from 1 setting
- Browser
- Not needed
Commerce
Listings, prices and sellers on marketplaces.
Every product a Shopify store publishes — title, price, compare-at price, stock, SKU, vendor, type, tags and images, a row per product, straight off the storefront’s own catalogue file with no account and no browser.
- You give it
- Stores
- You get back
- 22 columns, from 7 settings
- Browser
- Not needed
Jobs
Roles a company is hiring for, off the job boards and applicant trackers that publish them.
Indeed Jobs
Proxy advisedEvery employer advertising against a search on Indeed — role, company, their rating and review count, location, salary range, job type and how recently it was posted, a row per opening, with no account and no browser.
- You give it
- Searches
- You get back
- 27 columns, from 10 settings
- Browser
- Not needed
Every open role a company publishes on Greenhouse — title, location, department, office, when it was posted and when it last moved, a row per opening, off the board’s own public API with no account and no browser.
- You give it
- Boards
- You get back
- 16 columns, from 5 settings
- Browser
- Not needed
Every open role a company publishes on Ashby — title, team, location, workplace type, the published pay range and when it went up, a row per opening, off the board’s own public API with no account and no browser.
- You give it
- Boards
- You get back
- 21 columns, from 6 settings
- Browser
- Not needed
Every open role a company publishes on Lever — title, department, team, location, workplace type and commitment, a row per opening, off the board’s own public API with no account and no browser.
- You give it
- Boards
- You get back
- 15 columns, from 5 settings
- Browser
- Not needed
Running one
Open the Scrapers tab, pick a card, fill in its form and press Start. Every run started here makes a new table, named after what you searched for and when, filed in a folder named after the scraper — so two runs of the same card never overwrite each other. Check first if you like: it collects one page, writes nothing, and reports which columns came back filled.
You can also run one without opening the app at all. The assistant has the whole catalogue, so “collect the dentists in Rotterdam” is a sentence rather than a form; the same tools are on the MCP server for outside agents; and a Run scraper step puts a collector inside a workflow. That step is the one caller that can overrule the table-per-run rule above: left alone it files a new table like everything else, but pointed at a table you name it creates that one if it is missing and files every run into it after — appending, or updating the row in place on a column you match on. A scrape on a schedule then grows one table instead of leaving one behind a night. Either way the step hands the table’s name, its id and the row count to whatever comes next.
What they do not do
Almost nothing here signs in. Reading a public page signed out is the settled position — Bright Data scraped Meta signed out and won; hiQ scraped LinkedIn signed in and lost — so a card that needs an account moves a question from settled to open, and each one is a decision taken on its own rather than a default. LinkedIn is read entirely signed out here for that exact reason.
2 cards ask for an account, and only where the site leaves no signed-out source at all — not where signing in would simply return more:
- Instagram Profiles — Instagram no longer answers a signed-out request for a profile’s bio, follower counts or bio link — it returns a stub. This card needs an account of its own to read them. Everything else in Argus still collects signed out.
- Instagram Followers — A follower list is the one thing Instagram answers only to a signed-in request — signed out, every endpoint that serves one returns 401 “login required”, whatever address it is asked from. This card needs an account of its own to read them, and a throwaway one is the right kind. Everything else in Argus still collects signed out.
Even then no browser window opens to collect. One opens once, by hand, so you can log in; every request after that is the same plain HTTP every other card makes. Use a throwaway account — automated collection is against these sites’ terms and the account carrying it can be disabled.
What none of them will do is drive a page. Anything that needs a real browser — a site with no list endpoint behind it, or anything operating your own accounts — is an automation instead.
There is also no “scrape this URL for me”. A scraper is a card for a named service, written and measured against that service’s own replies. For a site no card covers, build an automation instead.