Threads Post Analyzer

Scrape posts from Threads (threads.net) by keyword and optionally enrich each post with Claude AI analysis — topic, sentiment, hook type, pain points, and a ready-to-use suggested hook for your own content.

Built for creators, marketers, and founders doing content research, audience discovery, or competitor analysis on Meta's Threads.

Features

• Search Threads by one or multiple keywords in a single run
• Extract posts directly from Threads' server-side data — stable, no fragile DOM selectors
• Collect per-post metrics: likes, replies, reposts, quotes
• Filter by time window (last N days)
• Optional AI enrichment via Claude (bring your own Anthropic key)
• Residential proxy out of the box (Apify RESIDENTIAL group)
• Pay-per-result pricing — you only pay for posts actually returned

Use Cases

• Content research — find what's resonating in your niche before you post
• Pain-point mining — discover what users complain about in your target audience
• Competitor analysis — see which hooks and topics your competitors' audience reacts to
• Trend tracking — monitor how a topic evolves on Threads week over week
• SDR / lead gen — spot founders and buyers publicly describing problems you solve

How to Use

1. Enter one or more keywords (e.g. indie hackers, solopreneur).
2. Set max posts per keyword (default 50).
3. Set the period in days (default 30 — older posts are skipped).
4. (Optional) Toggle AI analysis on and paste your Anthropic API key.
5. Run the Actor.
6. Download the results as JSON, CSV, Excel, or via API.

Output

Each item in the dataset contains:

Field	Description
postUrl	Direct link to the Threads post
author / authorHandle	Author display name and @username
text	Full post text
timestamp	ISO-8601 publish time
likes, replies, reposts, quotes	Engagement numbers
keyword	Which input keyword produced this post
ai.topic	2–4 word topic label (only if AI enabled)
ai.sentiment	positive / neutral / negative
ai.hookType	question / bold-claim / story / list / contrarian / other
ai.painPoints	Array of 1–5 pain points mentioned
ai.suggestedHook	One-line hook you can reuse for your own post

Example Input

{
    "keywords": ["indie hackers", "solopreneur"],
    "maxPosts": 100,
    "periodDays": 30,
    "enableAIAnalysis": true,
    "anthropicApiKey": "sk-ant-..."
}

Example Output Item

{
    "postUrl": "https://www.threads.net/@someuser/post/Abc123",
    "author": "Some User",
    "authorHandle": "someuser",
    "text": "Just hit $1K MRR after 6 months of building in public...",
    "timestamp": "2026-04-15T10:22:00Z",
    "likes": 412,
    "replies": 38,
    "reposts": 22,
    "quotes": 4,
    "keyword": "indie hackers",
    "ai": {
        "topic": "MRR milestone",
        "sentiment": "positive",
        "hookType": "story",
        "painPoints": ["slow growth in first months", "doubt before first revenue"],
        "suggestedHook": "What nobody tells you about the first 6 months of bootstrapping"
    }
}

Pricing

This Actor uses pay-per-result: $0.005 per post returned. You only pay for posts actually delivered. Failed requests and skipped items are free.

Platform infrastructure (proxy, compute) is included in the per-result price on the Apify side. If you enable AI analysis, you also pay Anthropic directly for Claude Haiku usage (roughly $0.0005 per post). The Apify side of the bill stays the same either way.

FAQ

Do I need my own Anthropic API key? Only if you enable AI analysis. You can get a key at console.anthropic.com. The Actor itself works fine without it — you just get posts without AI enrichment.

Does it work with private or age-restricted Threads? No. Only publicly visible posts are returned.

Why residential proxy? Threads aggressively rate-limits datacenter IPs. Residential proxy rotates per-request and drastically reduces blocks.

How fresh are the posts? Results reflect whatever the Threads search returns at run time — usually within minutes of publication.

Why are views not in the output? Threads only exposes view counts to the post author, not publicly. Other Actors that claim to return views are either guessing or scraping the author's own dashboard.

Support

Open an issue on the Actor's Apify Store page or reach the maintainer via the contact shown in the listing.