🧵 Threads.net Search & Conversation Scraper

A high-performance, resilient, and ultra-reliable B2B data scraper that extracts Meta Threads search results, posts, nested replies, profiles, and historical engagement into clean structured JSON/CSV datasets.

---

⚡ Why this Actor?

Most Threads scrapers break weekly because they rely on brittle DOM selectors or hardcoded GraphQL document hashes (doc_id). When Meta rotates these keys, competitor tools crash instantly — which is why their Apify Store ratings cluster around 1.2–2.8★.

This scraper uses a Playwright Network-Response Interception Engine:

1. Network Interception — Runs headless Playwright and hooks directly into live page network streams to capture official, encrypted /graphql and /ajax/ responses as the browser generates them organically. No hardcoded keys that go stale.
2. Recursive BFS Discussion Tree — Automatically follows conversation sub-replies recursively (Breadth-First Search) to map out the entire hierarchical discussion tree (max 5 levels deep).
3. Optional Cookie Injection — Inject a sessionid cookie to authenticate as a logged-in Threads user for higher rate limits and access to private-but-publicly-shared content. (100% cookieless by default.)
4. Three Modes — search (posts by keyword), bfs (recursive replies from a post URL), profile (posts + stats from a user timeline).

---

🪙 Monetization: Pay-Per-Result

• Actor Start Fee: $0.05 (covers setup cost)
• Result Fee: $0.0015 per result ($1.50 per 1,000 records)
• Bronze / Silver / Gold discounts available for Apify subscription users.
• Compute Costs: Free for users (compute is platform-paid).

---

📥 Input Parameters

Parameter	Type	Required	Default	Description
startUrls	Array	One of	[]	List of Threads URLs to scrape (Posts /post/CODE or User profiles @username).
search	String	One of	""	Keyword to search posts.
mode	String	No	"auto"	Mode: auto (auto-detect), search, bfs, profile.
maxItems	Integer	No	100	Maximum number of results to extract.
includeNestedReplies	Boolean	No	false	When enabled, recursively descends into comments-in-comments.
cookies	String	No	""	Optional sessionid=... cookie for authenticated scraping.
proxy	Object	Recommended	—	Apify Proxy config — Datacenter is fine; Residential for heavy use.

---

📊 Sample Output

Post Result

{
  "id": "18023456789012345",
  "item_type": "post",
  "url": "https://www.threads.net/@zuck/post/Dabc123",
  "author_username": "zuck",
  "author_display_name": "Mark Zuckerberg",
  "text_content": "Excited to share Threads has hit 200M monthly active users!",
  "created_at": "2026-07-12T08:00:30.000Z",
  "like_count": 12500,
  "reply_count": 832,
  "repost_count": 410,
  "view_count": 489000
}

Nested Reply (BFS)

{
  "id": "18023456789012345",
  "item_type": "reply",
  "url": "https://www.threads.net/@commenter/post/18023456789012345",
  "author_username": "commenter",
  "author_display_name": "John Doe",
  "text_content": "This is an extremely insightful thread!",
  "created_at": "2026-07-12T08:00:30.000Z",
  "like_count": 420,
  "reply_count": 8,
  "repost_count": 2,
  "source_post_url": "https://www.threads.net/@zuck/post/Dabc123",
  "parent_reply_id": "18023456789000111"
}

---

💡 Best Practices

• Use mode: 'auto' for mixed input — the Actor will detect post URLs vs profile URLs vs search keywords automatically.
• Inject cookies if you see "Login required" or 0-result responses — Threads sets per-IP rate limits aggressively.
• Enable includeNestedReplies only when you specifically need the full discussion tree — it's 3-5× more expensive per post.
• Use a Residential Proxy for high-volume runs (>500 results). Datacenter works for <100.

---

⚠️ Limitations

• Requires JavaScript execution (Playwright-based; can't be made HTTP-only).
• Meta rotates GraphQL signatures every 2-4 weeks; this Actor's auto-adapt will keep working without code updates.
• Deeply nested replies (depth > 5) are truncated to prevent runaway costs.