Threads Conversation Intelligence – Search, Replies & Monitor

Turn public Threads conversations into clean, monitor-ready data without a Threads login, browser session, or developer token.

Use one Actor for brand research, trend discovery, creator tracking, conversation analysis, customer-language research, and scheduled alerts. It combines several public search views, removes duplicates, preserves reply relationships where the public page exposes them, and reports partial coverage instead of quietly pretending a conversation is complete.

Start free

The default Free cached preview returns three representative records without contacting Threads and without a paid event. It lets you inspect the dataset shape before paying for a live query.

For live collection, select an operation and add your search terms, profiles, or post URLs.

What you can do

Search public conversations

• Search keywords, phrases, hashtags, products, competitors, or topics
• Merge Default, Recent, and Top search views
• Remove duplicate posts automatically
• Capture post text, author, timestamp, engagement, links, and media

Research public profiles

• Resolve public profile identity and biography
• Collect follower and verification information exposed publicly
• Return the profile's currently visible recent posts
• Distinguish unavailable or private targets from successful empty results

Analyze a post and its replies

• Return the root post and visible public replies
• Link direct replies to the root
• Infer deeper reply parents by the referenced public author when an exact parent post ID is not exposed
• Include a confidence label for every inferred link
• Compare the post's reported reply count with the number actually collected

Monitor only new results

Choose Only new results, select Search, Profile, or Conversation, and run the same Actor task on an Apify schedule. The Actor stores a bounded ID history and emits only records it has not seen before.

The first monitor run establishes the baseline and returns the currently visible records. Later runs return only additions.

Honest coverage

Threads often reports more replies than it exposes to a logged-out public page. Every conversation therefore includes a coverage record with:

• reportedReplyCount
• collectedReplyCount
• visibleCoveragePercent
• complete
• a plain-language coverage note

This makes incomplete public coverage measurable instead of invisible.

Pricing

The cached preview is free. Live use is charged per successful target, not per result row:

Live action	Price per target
Search term	$0.001
Public profile	$0.001
Post conversation	$0.003
Only-new monitor target	$0.001

A tiny Actor-start event covers the fixed cost of beginning a run; Apify currently calculates it at $0.00005 for this 256 MB Actor. A failed target is not charged. A search term can return many deduplicated records for one event. A normal one-target search therefore costs about $0.00105, while a conversation costs about $0.00305. Apify may display the equivalent per 1,000 events in its standard pricing table.

Input examples

Search a market topic

{
  "operation": "search",
  "searchTerms": ["electric vehicle", "EV charging"],
  "searchSurfaces": ["default", "recent", "top"],
  "maxResults": 50
}

Collect a public profile

{
  "operation": "profile",
  "profiles": ["@natgeo"],
  "maxResults": 30
}

Analyze visible replies

{
  "operation": "conversation",
  "postUrls": ["https://www.threads.com/@natgeo/post/Db5pTtZFAmB"],
  "maxReplies": 100
}

Publish a scheduled search monitor task

{
  "operation": "monitor",
  "monitorType": "search",
  "searchTerms": ["your brand", "your product"],
  "searchSurfaces": ["recent"],
  "maxResults": 100
}

Output

Rows use one of three types:

• profile — public identity, biography, follower information, and verification
• post — content, author, engagement, media, permalink, reply relationship, and timestamps
• coverage — reported versus collected conversation depth

Every live row also carries queryType, queryTarget, and monitorRun, so results from multi-target runs remain easy to filter.

The run's OUTPUT record contains target success/failure counts, retry attempts, response sizes, emitted rows, billing count, and errors. One broken target does not discard successful targets from the same run.

Reliability design

• Reads server-rendered public pages instead of fragile visual selectors
• Validates that a response contains real structured Threads data rather than a login shell
• Retries timeouts, rate limits, server errors, and incomplete responses
• Uses conservative bounded concurrency
• Uses Apify Proxy by default in the cloud and rotates routing on retries; API users can disable it explicitly
• Keeps successful search views when another selected view is temporarily rate-limited
• Deduplicates across search views and repeated embedded payloads
• Keeps per-target failures visible in the run summary

Public-data boundary

This Actor reads only information exposed on public Threads pages. It does not log in, access private accounts, solve CAPTCHAs, send messages, like content, or automate user accounts.

Public availability does not replace your own legal obligations. Use the Actor only where your purpose, collection, storage, and processing comply with applicable law, the rights of data subjects, and the relevant platform terms. Site behavior and public coverage can change.

Threads and Meta are trademarks of their respective owners. This Actor is an independent tool and is not affiliated with or endorsed by Meta.

Support

When reporting a problem, include the public target, run ID, operation, and whether the failure affected every target or only one. Do not send login cookies, access tokens, or private content.