Instagram Comments Scraper Pro

Comment intelligence from Instagram posts: sentiment analysis, spam detection, engagement scoring, and MCP-ready providerHealth metadata. Desktop+embed fallback chain. No login cookies required.

---

Why This Actor Is Better

Competitor comparison

Feature	This Actor	Apify IG Comments (top)	Free IG scrapers	Sentiment SaaS
No login cookies required	✅	❌	❌	n/a
Desktop+embed fallback	✅	❌	❌	n/a
Keyword sentiment	✅ open-source	❌	❌	✅ paid
Spam detection	✅ heuristics	❌	❌	✅ paid
Engagement labels	✅	❌	❌	partial
Optional OpenAI NLP	✅ user key	❌	❌	✅ paid
Confidence score (0-1)	✅	❌	❌	❌
MCP-ready metadata	✅ providerHealth	❌	❌	❌
Price / 1K comments	$1.77	~$2.50	free (rate-limited)	~$5-20

Key Features

• 🛡️ Multi-API fallback chain — Instagram desktop (Playwright) primary with automatic embed/oembed HTML fallback.
• 💬 Sentiment analysis — open-source keyword classifier (positive/negative/neutral) on every comment.
• 🚫 Spam detection — heuristic scoring flags spam/suspicious/clean comments (URLs, keywords, ALL CAPS).
• 📊 Engagement labels — high/medium/low/none based on likes.
• 🎯 Confidence scoring — 0.0–1.0 reliability score on every result.
• 🔗 Source attribution — know which providers contributed each field.
• ⚡ Cache-first mode — fast_lookup hits KVS cache (30min TTL — comments change fast).
• 🤖 MCP-ready — providerHealth{} on every result for agent routing.
• 🔌 Optional OpenAI NLP — drop in OPENAI_API_KEY for GPT-4o-mini sentiment. Disabled by default.
• 💰 Transparent PPE pricing — pay only for successful comments.

---

Architecture

flowchart TD
    A[Input: post URLs + mode] --> B{Cache hit?}
    B -- yes --> C[Return cached base comments]
    B -- no --> D[Primary: Instagram desktop Playwright]
    D -- fails --> E[Fallback: Instagram embed HTML httpx]
    D --> F[Normalize: author/text/likes/timestamp]
    E --> F
    F --> G[Enrichment layer]
    G --> G1[Keyword sentiment]
    G --> G2[Spam detection heuristics]
    G --> G3[Email/phone/URL extraction]
    G --> G4[Engagement labels]
    G --> G5[Optional: OpenAI NLP sentiment]
    G1 --> H[Confidence scoring + source attribution]
    G2 --> H
    G3 --> H
    G4 --> H
    G5 --> H
    H --> I[Progressive dataset push]
    I --> J[Webhook + MCP-ready metadata]
    C --> J

---

Modes

Mode	Description	Target latency	Use case
fast_lookup	Cache-first, base comments only	<800ms cached	Quick counts, dedup
enrich	Sentiment + spam + engagement + optional NLP	~2-4s/post	Brand monitoring, audience analysis
batch	Queue-based, full enrichment, per-item isolation	varies	Large post lists (100+)

---

Input

Parameter	Type	Required	Default	Description
mode	string	—	enrich	fast_lookup | enrich | batch
queries	array	✅	["https://instagram.com/p/Cxxx"]	Instagram post URLs
maxResults	integer	—	25	Max comments per post (1–1000)
webhookUrl	string	—	—	Webhook for completion notification

Example input

{
  "mode": "enrich",
  "queries": ["https://www.instagram.com/p/Cxxx/", "https://www.instagram.com/p/Cyyy/"],
  "maxResults": 100,
  "webhookUrl": "https://your-app.com/webhook"
}

---

Output

Field	Type	Description
query	string	Input query
postUrl	string	Source post URL
author	string	Commenter handle
text	string	Comment text
timestamp	string	Comment timestamp
likes	integer	Like count
engagementLabel	string	high | medium | low | none
sentiment	string	positive | negative | neutral
sentimentNlp	string | null	OpenAI NLP sentiment (if key provided)
spamLabel	string	clean | suspicious | spam
spamScore	number	0.0–1.0 spam probability
emails	array	Emails found in comment
phones	array	Phones found in comment
urls	array	URLs found in comment
confidenceScore	number	0.0–1.0 reliability
sources	array	Provider attribution
providerHealth	object	Per-provider status/latency
cacheStatus	string	hit | miss | degraded
mode	string	Execution mode
extractedAt	string	ISO timestamp

Example output

{
  "query": "https://www.instagram.com/p/Cxxx/",
  "postUrl": "https://www.instagram.com/p/Cxxx/",
  "author": "jane_doe",
  "text": "Love this! So beautiful 😍",
  "timestamp": "2026-06-28T10:00:00Z",
  "likes": 42,
  "engagementLabel": "medium",
  "sentiment": "positive",
  "spamLabel": "clean",
  "spamScore": 0.0,
  "emails": [],
  "phones": [],
  "urls": [],
  "confidenceScore": 0.8,
  "sources": ["instagram", "sentiment_analyzer", "spam_detector"],
  "providerHealth": {
    "instagram_desktop": {"status": "ok", "latency_ms": 3200, "error": null},
    "sentiment_analyzer": {"status": "ok", "latency_ms": 0, "error": null},
    "spam_detector": {"status": "ok", "latency_ms": 0, "error": null}
  },
  "cacheStatus": "miss",
  "mode": "enrich",
  "extractedAt": "2026-06-28T23:50:00.000Z"
}

---

Pricing

Plan	Price per 1K comments	Savings vs. top competitor
Leading competitors	~$2.50/1K	—
This actor (≤10K/mo)	$1.77/1K	29% cheaper
This actor (10K–100K/mo)	$1.50/1K	40% cheaper
This actor (100K+/mo)	$1.25/1K	50% cheaper

Optional event: sentiment_scored at $0.30/1K comments with sentiment analysis.

---

Use Cases

• Brand monitoring — track sentiment across product launch posts
• Audience analysis — identify engaged commenters and filter spam
• Crisis detection — spike in negative sentiment alerts PR teams
• Influencer vetting — measure real engagement vs spam in comment sections
• Lead generation — extract emails/phones left in comments
• MCP agent workflows — providerHealth lets agents route around failures
• Content research — find high-engagement comments for testimonials
• Compliance — flag spam/promotional/policy-violating comments at scale

---

Integration Examples

Python (Apify SDK)

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("ayeeyee/instagram-comments-scraper-pro").call(run_input={
    "mode": "enrich",
    "queries": ["https://www.instagram.com/p/Cxxx/"],
    "maxResults": 100,
})

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"@{item['author']}: {item['sentiment']} | spam={item['spamLabel']} | {item['text'][:50]}")

cURL

curl -X POST "https://api.apify.com/v2/acts/ayeeyee~instagram-comments-scraper-pro/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"mode":"enrich","queries":["https://www.instagram.com/p/Cxxx/"],"maxResults":50}'

MCP (Model Context Protocol)

$npx -y @apify/actors-mcp-server --tools actors,ayeeyee/instagram-comments-scraper-pro

Agents can call call-actor and use providerHealth + spamLabel + sentiment to filter and route.

---

FAQ

Q: Do I need Instagram login cookies? No. Public post comments are scraped via DOM selectors and meta tags — no login required. Private posts return no comments.

Q: How accurate is the sentiment? The default keyword classifier is ~70-80% accurate for obvious positive/negative comments. For higher accuracy, set OPENAI_API_KEY to enable GPT-4o-mini NLP sentiment.

Q: What does the spam detector flag? Comments with spam keywords ("dm me", "free follower", "buy now"), multiple URLs, all-caps shouting, or very short comments with links. Score 0.5+ = spam, 0.3-0.5 = suspicious, <0.3 = clean.

Q: Can I call this from an LLM agent? Yes. MCP-ready with providerHealth{}, sentiment, spamLabel, and confidenceScore for routing/filtering decisions.

Q: What's the cache TTL? 30 minutes — comments change faster than profiles or business listings.

---

Legal & Compliance

Scrapes publicly available Instagram comment data. Does not access private data, bypass authentication, or store credentials. Users are responsible for complying with GDPR/CCPA and Instagram's ToS.

---

AI-DLC / Data Lifecycle

• Collection — Public data only; respects robots.txt and rate limits.
• Processing — In-memory normalization; no PII logging.
• Storage — Results in user's Apify dataset, not retained by actor.
• Usage — Brand monitoring, audience analysis, legitimate research.
• Disposal — No long-term caching (30min TTL for base results only).

---

Enhancement Roadmap (API / MCP Integrations)

• OpenAI sentiment MCP — GPT-4o-mini NLP sentiment (optional, user key)
• Google Perspective API MCP — toxicity scoring
• LangGraph workflow — IG comments → sentiment aggregation → alert routing
• Vector store — semantic comment clustering across posts

---

Changelog

• v3.0 — Multi-API orchestration edition: desktop+embed fallback, spam detection, engagement labels, MCP-ready providerHealth, optional OpenAI NLP, expanded FAQ, integration examples, volume pricing.
• v2.0 — Premium README, AI-DLC docs, confidence scoring, source attribution.
• v1.0 — Initial release with Playwright scraping and keyword sentiment.

---

Links

• Apify Store: https://apify.com/ayeeyee/instagram-comments-scraper-pro
• Actor ID: Jgfr5cR4ctbyPbX79