📺 YouTube Comments Scraper

Scrape YouTube comments and replies from any video URL or ID using the official YouTube Data API v3. Returns text, author, likes, timestamps, language, and channel info — JSON or CSV.

🎯 Built for sentiment analysis, AI/LLM training datasets, brand monitoring, creator research, and audience insights.

---

✨ What you can do

• 🎥 Multiple videos per run — paste any list of URLs or IDs
• 💬 Top-level comments + reply threads — flattened with parentId for tree reconstruction
• 🔎 Optional substring filter — only return comments containing your keywords
• 📊 Sort by relevance (top) or time (newest)
• 📺 Includes video metadata — title, channel, view/like/comment counts

---

🚀 Quick start

{
    "videoUrlsOrIds": [
        "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
        "kffacxfA7G4"
    ],
    "apiKey": "AIzaSy...",
    "includeReplies": true,
    "order": "relevance",
    "maxCommentsPerVideo": 500,
    "searchTerms": ["amazing", "great"]
}

---

🔑 Get the API key (free, 5 min)

1. Open https://console.cloud.google.com/
2. Create a project (any name)
3. Search "YouTube Data API v3" → click → Enable
4. Left sidebar: APIs & Services → Credentials
5. Click "+ Create credentials" → "API key"
6. Recommended: Restrict key → API restrictions → YouTube Data API v3
7. Paste the AIzaSy... key into the input form

Free quota: 10,000 units/day per project.

• 1 page of comments (~100) = 1 unit
• Average video (500 comments) = ~5 units
• 2,000 average videos/day for free

---

📥 Input

Field	Description
videoUrlsOrIds	List of YouTube URLs or 11-char video IDs
apiKey	Required — your YouTube Data API v3 key
includeReplies	Fetch reply threads (default true)
order	relevance (top) or time (newest)
maxCommentsPerVideo	Cap on top-level comments per video
searchTerms	Optional substring filter

URL formats supported: youtube.com/watch?v=ID, youtu.be/ID, youtube.com/shorts/ID, youtube.com/embed/ID.

---

📤 Output (per comment)

{
    "videoId": "dQw4w9WgXcQ",
    "videoTitle": "Rick Astley - Never Gonna Give You Up",
    "channelTitle": "Rick Astley",
    "commentId": "Ugw...",
    "parentId": null,
    "isReply": false,
    "author": "@username",
    "authorChannelId": "UC123...",
    "authorProfileImage": "https://...",
    "text": "Comment text",
    "textOriginal": "Comment text",
    "likeCount": 234,
    "publishedAt": "2026-04-15T...",
    "updatedAt": "2026-04-15T...",
    "totalReplyCount": 5,
    "scrapedAt": "2026-05-06T..."
}

---

🎯 Use cases

Who	Why
🤖 AI / sentiment teams	Real audience reactions for training/benchmarking
🎨 Creators & marketers	Audience insights, content optimization
🏢 Brand teams	Track mentions on competitors' / influencers' videos
📊 Researchers	Study online discourse around news/topics/products

---

⚙️ Tech notes

• Uses YouTube Data API v3 — official, terms-compliant
• Each user provides their own API key (we never store it)
• Auto-handles videos with disabled comments — logs and skips, doesn't fail
• Pagination via nextPageToken until cap or end

---

❓ FAQ

Will it use up my quota? A typical video scrape costs ~5 quota units. 10,000/day = ~2,000 typical scrapes.

What about comment replies? We fetch them automatically when includeReplies is on. They appear as separate items with isReply: true and parentId set.

Hidden / spam comments? The API only returns publicly visible comments. Hidden ones aren't included.