Instagram Keyword Scraper

Start from a keyword, not a username.

Most Instagram scrapers need you to already know the account. This one takes a topic — a name, a niche, a hashtag you half-remember — tries the spellings people actually use, and comes back with public post and reel candidates plus a summary of what worked.

How it searches

Give it kim chaeyeon and it tries, on its own:

Variant	Example
exact keyword	kim chaeyeon
compact hashtag	#kimchaeyeon
underscore hashtag	#kim_chaeyeon
profile guess	@kimchaeyeon, @kim_chaeyeon
keyword search	Instagram's own results page

If a niche needs more, add aliases after the first run — romanized names, stage names, brand handles — and the same pass runs across all of them.

What you get

Rows are tagged by type, so one filter separates them:

recordType	One row per
search	keyword — what was tried and how it went
searchCandidate	candidate page found for that keyword
media	resolved public post or reel
runSummary	the run itself — totals, usable rows, failures, shortages

The runSummary row is the one to read first: total rows, usable media rows, failed rows, target shortages, and the candidates that produced the most.

Input

Three fields, nothing else to tune.

{
  "keywords": "karina\nkim chaeyeon\nseoul food",
  "resultSize": "scout25",
  "keywordAliases": ""
}

Field	Meaning
keywords	One per line
resultSize	scout25 cheap first check · scout100 standard batch · advanced300 deeper research run
keywordAliases	Optional. karina: aespakarina, karinaaespa — one keyword per line

Start with scout25 and no aliases. Read the summary, then add aliases only for the keywords that came back thin.

What it is good for

• Creator and model scouting in a niche, before manual review
• UGC and brand-mention discovery from product or campaign words
• Local trend research — food, travel, venues, events
• Alias testing — which spelling of a name actually surfaces content

Honest limits

• This is discovery, not full coverage. It finds public candidates from keywords. It is not a replacement for a broad profile scraper and does not promise every post for a term.
• Public pages only. No login, no private accounts, no follower exports.
• Availability moves. Instagram changes what public pages serve; a keyword that was rich last week can be thin today. That is why every run ends with a summary instead of a silent short count.
• A measured scout25 run: 19 rows, 15 usable media rows, 0 failures, about 2.4 minutes. Treat it as a smoke-test expectation, not a guarantee.

More actors from this author

• YouTube Channel Search — find channels and creators by keyword
• YouTube Channel Scraper — one channel to profile, every video and the comments
• YouTube Search Scraper — keyword search to videos with exact views and likes
• Google Maps Reviews Scraper — reviews in the reviewer's own language
• Naver Shopping Scraper — Korean product search, prices and sellers

Questions and problems

Open an issue on the Issues tab of this actor, with the run ID.