🔍 Instagram Keyword Search Scraper — Posts & Reels

🔗 View on Apify Store | 🇺🇸 English | 🇧🇷 Português

Search Instagram by keyword or hashtag and extract matching posts, Reels, and carousels. Get permalinks, captions, authors, media URLs, hashtags, mentions, and engagement when Instagram exposes them. Playwright cascade: cookie seeding → keyword explore XHR → hashtag page → embedded JSON → DOM links. No Meta API key.

✨ Features

• 🔍 Keyword & hashtag search — Single-word (travel) and multi-word (sunset photography)
• 📝 Posts, Reels & Carousels — Discover content from Instagram search, not from a profile you already know
• 📊 Engagement when exposed — Likes, comments, video views, duration
• 🏷️ Hashtags & mentions — Parsed from the caption
• 🧵 Multiple keywords — Process a batch in one run
• 🍪 Cookie support — Optional session cookies when Instagram shows a login wall
• 🔄 Strategy cascade — Explore XHR → hashtag page → embedded JSON → DOM permalinks
• 🛡️ Stealth browsing — Webdriver flag removal and basic fingerprint spoofing
• 🌐 Proxy control — Datacenter default; residential only after a confirmed block
• ✅ Contract + provenance — Invalid items are dropped, never billed
• 📷 Debug screenshots — Saved to KV Store when a keyword yields zero posts

🧭 Capabilities & Limits

Discover by	Supported?	Notes
Keyword / search query (keywords)	✅ Yes	Primary input. Returns posts/Reels/carousels.
Hashtag (#tag as a keyword)	✅ Yes	Also opens /explore/tags/{tag}/ as fallback.
Handle / username (profiles)	❌ No	This actor searches by keyword. Use the Reels & Posts Scraper for profiles.
Post / Reel URL	❌ No	Search results only.
Accounts / places typeahead	❌ No	v1.0 typeahead was replaced. This actor returns posts.

Output includes: post URL ✅ · shortcode ✅ · type ✅ · caption ✅ (when the media node has it) · author ✅ (when exposed) · likes/comments/views ✅ (XHR/embedded) or null on DOM fallback · hashtags/mentions ✅ (from caption) · comment text ❌.

📥 Input

Parameter	Type	Required	Default	Description
keywords	string[]	Yes	—	Keywords or hashtags (sunset photography, #fitness)
maxPosts	integer	No	10	Max posts per keyword (1–100 in this version)
maxResults	integer	No	—	Alias of maxPosts (unified vocabulary)
seeds	string[]	No	—	Alias of keywords when seedType is keyword
seedType	string	No	keyword	Only keyword is supported
cookies	object[]	No	—	Optional {name, value, domain} session cookies
proxyConfig	proxy	No	Apify DC	Datacenter default

Input Example

{
  "keywords": ["sunset photography", "#fitness"],
  "maxPosts": 10
}

With cookies for authenticated access:

{
  "keywords": ["sunset photography"],
  "maxPosts": 10,
  "cookies": [
    {
      "name": "sessionid",
      "value": "your-session-id",
      "domain": ".instagram.com"
    }
  ]
}

📤 Output

Field	Type	Description
keyword	string	Keyword that produced this item
id	string | null	Instagram media id when exposed
shortCode	string	Shortcode used in the permalink
url	string	https://www.instagram.com/p/{code}/ or /reel/
type	string	post, reel, or carousel
caption	string	Caption text when present
author	string | null	Username when the media node includes it
timestamp	string | null	ISO 8601 UTC publish time
likesCount	integer | null	Likes when exposed
commentsCount	integer | null	Comment count (no comment text)
videoViewCount	integer | null	Plays/views for video
videoDuration	number | null	Duration in seconds
displayUrl	string | null	Thumbnail / display image
hashtags	string[]	Parsed from caption
mentions	string[]	Parsed from caption
source	string	xhr, embedded, or dom
_dataQuality	string	full, dom_only_no_metrics, or diagnostic
scrapedAt	string	ISO 8601 UTC collection time

Missing engagement is null, never invented. DOM-only items have permalink + shortcode and _dataQuality: "dom_only_no_metrics".

Output Example

{
  "keyword": "sunset photography",
  "id": "3315678901234567890",
  "shortCode": "DFxRtPqWsYz",
  "url": "https://www.instagram.com/p/DFxRtPqWsYz/",
  "type": "post",
  "caption": "Golden hour in Lisbon #travel #sunset",
  "author": "citywalks",
  "timestamp": "2026-08-10T19:12:00Z",
  "likesCount": 1842,
  "commentsCount": 36,
  "videoViewCount": null,
  "videoDuration": null,
  "displayUrl": "https://scontent.cdninstagram.com/v/t51.2885-15/428573961_1234567890.jpg",
  "hashtags": ["travel", "sunset"],
  "mentions": [],
  "source": "xhr",
  "_dataQuality": "full",
  "scrapedAt": "2026-08-16T20:00:00Z"
}

📋 Use Cases

• 🔎 Trend monitoring — See what Instagram surfaces for a topic or hashtag
• 🏷️ Hashtag research — Collect real posts behind a tag, not just the tag page title
• 📊 Market research — Caption language, hashtags, and engagement patterns by keyword
• 🤖 Seed generation — Feed authors into the Reels & Posts Scraper for a full profile pull
• 📈 Campaign tracking — Watch branded keywords over time

❓ FAQ

Q: Does this use the official Instagram Graph API? A: No. Graph API does not offer anonymous keyword search of public posts. This actor reads Instagram's public web search UI.

Q: Does it work without logging in? A: Often yes, via homepage cookie seeding. A login wall is common. If you get _blockReason: BLOCKED, provide cookies (sessionid).

Q: Why are likes null on some rows? A: DOM fallback found a real permalink but the media JSON did not include metrics. _dataQuality is dom_only_no_metrics. Those rows are still real posts.

Q: Why is maxPosts capped at 100? A: Cost control. This is Playwright. A 10,000-post crawl is not safe until we have a measured cost curve.

Q: Who pays compute and proxy? A: You do. Pricing is Pay per event + usage. Events cover each valid post; compute units and Apify proxy are billed to the run owner. The publisher does not absorb Playwright or residential cost.

Q: What happens on zero posts? A: Confirmed login wall or an empty set from Instagram → SUCCEEDED + one diagnostic item, no PPE. Parser drift or no origin response → FAIL with status, URL, HTML size, intercept count and proxy flag.

💰 Pricing

Pay Per Event + platform usage (client pays CU, traffic and Apify proxy):

Event	Price
apify-actor-start	$0.00005 (platform default; 1× per GB of memory)
post-discovered	$0.001 per valid post ($1.00 / 1,000)

Example: 5 keywords × 10 posts = 50 items = $0.05 in events, plus the run's compute/proxy on your Apify invoice.

Diagnostics are never charged. Free-plan users are capped at 5 posts per keyword.

🔗 Related Actors

• Instagram Reels & Posts Scraper — Profile grid + full engagement
• TikTok Viral Scanner — TikTok profile and video data
• YouTube Fast Scraper — YouTube video metrics
• Google Maps BR Scraper — Google Maps business data

📝 Changelog

v1.1.0 (Current)

• Product is posts/Reels by keyword, not typeahead users/hashtags/places
• PPE event post-discovered; incremental charge; owner-skip
• Platform usage paid by the client (isPPEPlatformUsagePaidByUser)
• UAG-SG only on confirmed block or empty origin
• Residential fallback only after a confirmed block
• Default run: 2048 MB / 600 s

v1.0

• Typeahead search (users / hashtags / places) — superseded

---

🔍 Instagram Keyword Search Scraper — Posts e Reels

🇺🇸 English | 🇧🇷 Português

Pesquise por keyword ou hashtag no Instagram e extraia posts, Reels e carrosséis. Permalink, legenda, autor, URL de mídia, hashtags, menções e engajamento quando o Instagram expõe. Cascata Playwright: cookie seeding → explore XHR → página de hashtag → JSON embutido → links DOM. Sem chave da API da Meta.

✨ Funcionalidades

• 🔍 Busca por keyword e hashtag — Uma palavra (viagem) ou várias (pôr do sol lisboa)
• 📝 Posts, Reels e carrosséis — Conteúdo da busca, não de um perfil que você já conhece
• 📊 Engajamento quando exposto — Curtidas, comentários, views, duração
• 🏷️ Hashtags e menções — Extraídas da legenda
• 🧵 Várias keywords — Um único run
• 🍪 Cookies opcionais — Quando o Instagram pede login
• 🔄 Cascata de estratégias — XHR da busca → hashtag → JSON embutido → permalinks DOM
• 🛡️ Navegação stealth — Remoção da flag webdriver
• 🌐 Proxy — Datacenter padrão; residencial só após bloqueio confirmado
• ✅ Contrato + proveniência — Item inválido é descartado, nunca cobrado
• 📷 Screenshots de debug — Salvos no KV Store se a keyword não render posts

🧭 Capacidades e Limites

Descobrir por	Suportado?	Observações
Keyword / busca (keywords)	✅ Sim	Entrada principal. Devolve posts/Reels/carrosséis.
Hashtag (#tag como keyword)	✅ Sim	Também abre /explore/tags/{tag}/ como fallback.
Handle / username (profiles)	❌ Não	Este actor busca por keyword. Use o Reels & Posts Scraper para perfis.
URL de post / Reel	❌ Não	Só resultados de busca.
Autocomplete de contas/lugares	❌ Não	O v1.0 typeahead foi substituído. Este actor devolve posts.

A saída inclui: URL do post ✅ · shortcode ✅ · tipo ✅ · legenda ✅ (quando o nó de mídia tem) · autor ✅ (quando exposto) · likes/comentários/views ✅ (XHR/embutido) ou null no fallback DOM · hashtags/menções ✅ · texto de comentário ❌.

📥 Entrada

Parâmetro	Tipo	Obrigatório	Padrão	Descrição
keywords	string[]	Sim	—	Keywords ou hashtags
maxPosts	inteiro	Não	10	Máximo de posts por keyword (1–100)
maxResults	inteiro	Não	—	Alias de maxPosts
cookies	objeto[]	Não	—	Cookies de sessão opcionais
proxyConfig	proxy	Não	DC Apify	Datacenter padrão

Exemplo de Entrada

{
  "keywords": ["sunset photography", "#fitness"],
  "maxPosts": 10
}

📤 Saída

Campos iguais à tabela em inglês. Itens só-DOM têm permalink + shortcode e _dataQuality: "dom_only_no_metrics". Engajamento ausente vem null.

📋 Casos de Uso

• 🔎 Monitoramento de tendência — O que o Instagram mostra para um tema
• 🏷️ Pesquisa de hashtags — Posts reais atrás da tag
• 📊 Pesquisa de mercado — Linguagem de caption e padrões de engajamento
• 🤖 Geração de sementes — Autores para o scraper de perfil
• 📈 Rastreamento de campanha — Keywords de marca ao longo do tempo

❓ Perguntas Frequentes

P: Usa a Graph API oficial? R: Não. A Graph API não oferece busca anônima de posts por keyword.

P: Funciona sem login? R: Muitas vezes sim, com cookie seeding da homepage. Login wall é comum — use cookies.

P: Por que likes vêm null? R: Fallback DOM achou permalink real sem JSON de métricas. _dataQuality = dom_only_no_metrics.

P: Por que o teto é 100? R: Custo de Playwright. Sem curva medida, 10 mil posts não é seguro.

P: Quem paga compute e proxy? R: Você. Pay per event + usage. Eventos cobrem cada post válido; CU e proxy Apify vão na conta de quem rodou.

P: E se não vier nenhum post? R: Login wall ou conjunto vazio confirmado pela origem → SUCCEEDED + diagnóstico, sem PPE. Drift ou origem muda → FAIL com status, URL, bytes e interceptações.

💰 Preços

Pay Per Event + uso de plataforma (o cliente paga CU, tráfego e proxy Apify):

Evento	Preço
apify-actor-start	$0.00005 (default; 1× por GB)
post-discovered	$0.001 por post válido ($1,00 / 1.000)

Exemplo: 5 keywords × 10 posts = 50 itens = $0,05 em eventos, mais o compute/proxy na sua fatura Apify.

Diagnóstico nunca é cobrado. Plano free: 5 posts por keyword.

🔗 Actors Relacionados

• Instagram Reels & Posts Scraper — Grade de perfil + engajamento
• TikTok Viral Scanner — Perfis e vídeos do TikTok
• YouTube Fast Scraper — Métricas do YouTube
• Google Maps BR Scraper — Empresas no Google Maps

📝 Changelog

v1.1.0 (Atual)

• Produto = posts/Reels por keyword
• Evento post-discovered; charge incremental; owner-skip
• Uso de plataforma pago pelo cliente
• UAG-SG só com bloqueio ou origem vazia confirmados
• Residencial só após bloqueio confirmado
• Default: 2048 MB / 600 s

v1.0

• Busca typeahead (users / hashtags / places) — substituída