GMGN Crypto KOL Twitter & Telegram Signal Feed — Real-time Influencer & Channel Calls

The fastest alpha feed on Apify. This actor pulls GMGN.ai's curated KOL Twitter activity (tweets, reposts, quotes, replies, follows, banner / handle / name / description changes) AND Telegram channel signals — all tagged, structured, and ready to ingest into your bot / dashboard / newsletter.

Every event comes pre-classified by user category (KOL, trader, exchange, celebrity, politics, media, founder, Binance Square, exchange-listing), with full content + auto-translation, source tweet, follower count, and (when relevant) the token address being discussed.

---

🚀 What this actor does

GMGN.ai monitors thousands of curated crypto Twitter accounts and Telegram channels in real time. Their backend tags every event with user_tags (account category) and the token being referenced (parsed from cashtags, contract addresses, etc.). This actor exposes the two raw feeds:

• /vas/api/v1/twitter/messages — Twitter activity feed with rich filtering.
• /vas/api/v1/tg/messages — Telegram channel signals (each tagged with win-rate, member count, mcap, current price).

Run it once → get the most recent snapshot. Run it with pollIterations: 120 and pollIntervalSec: 30 → 60-minute continuous live feed within a single Apify run.

---

💡 Use cases

• Alpha-feed Telegram / Discord bot — webhook every KOL tweet that mentions a token into your private signal channel.
• Sentiment tracking — collect KOL chatter over weeks; map mentions to price action.
• AI training — labelled dataset of KOL posts × token outcomes for predicting price impact.
• Influencer accountability — track when public figures pump / dump tokens.
• Listing watch — user_tags: ["exchange", "binance_square", "exchange_listing"] to be first to know about listings.
• News dashboard — surface real-time KOL reactions to events.
• Channel discovery — Telegram feed surfaces channel win_rate and members_count — ideal for vetting alpha channels.
• Anti-pump-and-dump detection — when an account renames, deletes posts, or changes handle, GMGN exposes those events too.

---

⚙️ Input configuration

Field	Type	Default	Description
feeds	string[]	["twitter","telegram"]	Pick twitter, telegram, or both.
userTags	string[]	curated set	Twitter user categories. Leave empty for all.
twTypes	string[]	["tweet","repost","quote","reply","pin"]	Twitter event types.
hasToken	boolean	false	Keep only Twitter posts with a parsed token contract.
featuredOnly	boolean	false	GMGN curated featured items only.
mineOnly	boolean	false	Try mine=1 (needs auth — generally leave OFF).
tgFeatured	boolean	false	Telegram featured-only.
tgLimit	integer	30	Telegram messages per call (max 30).
pollIterations	integer	1	How many polls within a single run.
pollIntervalSec	integer	30	Seconds between polls (when iterations > 1).
deduplicate	boolean	true	Skip items already seen this run.
useProxy	boolean	true
proxyConfiguration	object	residential
maxConcurrency	integer	2

Twitter user_tags vocabulary

Tag	Who it covers
kol	Key opinion leader.
trader	Pro trader / fund.
master	Crypto OG / master.
founder	Project founder / team.
exchange	Exchange official accounts.
celebrity	Mainstream celebrity.
politics	Political figures.
media	Crypto media outlets.
companies	Listed crypto companies.
binance_square	Binance Square posts.
exchange_listing	Listing-announcement accounts.
other	Other curated.

Twitter tw_types vocabulary

Type	Meaning
tweet	Original tweet.
repost	Retweet.
quote	Quote-tweet.
reply	Reply.
delete_post	Tweet deletion event.
pin / unpin	Pinning a tweet.
follow / unfollow	Following events (rare).
banner / photo	Profile media change.
name / handle / description	Profile-text change.

---

📦 Output rows

Two row kinds distinguished by _kind.

twitter

Field	Description
platform	twitter.
twType	Event type.
tweetId, sourceId	Tweet IDs.
userTwitterId, userScreenName, userName, userAvatar, userFollowers, userVerified	Author.
userTags	GMGN tag list.
sourceUserTwitterId, sourceUserScreenName, sourceUserName, sourceUserFollowers	Source account (for reposts / quotes).
sourceContent	Raw tweet text.
translationZh	Auto-translation to Chinese.
chain, tokenAddress, tokenSymbol, tokenLogo, price, mcap	Token tagged (when applicable).
twTimestamp, twTimestampMs	When the tweet was posted.
tweetUrl, userProfileUrl	Convenience links.
scrapedAt

telegram

Field	Description
platform	telegram.
channelId, channelName, channelUsername, channelAvatar, channelUrl	Channel identity.
channelMembersCount, channelWinRate	Channel metadata.
messageId, messageSendAt, messageSendAtUnix	Message identity & time.
content	Message content.
chain, tokenAddress, tokenSymbol, tokenLogo, price, price24h, mcap	Token tagged.
featured	Featured flag.
scrapedAt

---

🧪 Example output (twitter)

{
  "_kind": "twitter",
  "platform": "twitter",
  "twType": "repost",
  "tweetId": "2055258017877090314",
  "userScreenName": "MindMatterMoney",
  "userName": "Leon",
  "userFollowers": 261424,
  "userVerified": false,
  "userTags": ["kol"],
  "sourceUserScreenName": "LumenX_",
  "sourceContent": "\"I will give you back seven times what you lost.\" Joel 2:25.\n\nIf you believe type 'Amen'.",
  "translationZh": "「我必七倍偿还你所失去的。」约珥书 2:25\n\n如果你相信，请回复\"阿门\"。",
  "twTimestamp": "2026-05-15T12:04:30.966Z",
  "tweetUrl": "https://x.com/MindMatterMoney/status/2055258017877090314",
  "userProfileUrl": "https://x.com/MindMatterMoney",
  "scrapedAt": "2026-05-15T12:05:00.000Z"
}

🧪 Example output (telegram)

{
  "_kind": "telegram",
  "platform": "telegram",
  "channelId": "-1001664356761",
  "channelName": "Rich Duck Calls",
  "channelUsername": "richduckofficial",
  "channelMembersCount": 3966,
  "channelWinRate": 0.42,
  "channelUrl": "https://t.me/richduckofficial/19653",
  "messageId": "20607664128",
  "messageSendAt": "2026-05-15T12:02:20.000Z",
  "content": "$DIAMOND\n\n85k\n\n8Kpqq6791DvdomkV59DXMqRXvoh3k6wM6nbFNkcBpump",
  "chain": "sol",
  "tokenAddress": "8Kpqq6791DvdomkV59DXMqRXvoh3k6wM6nbFNkcBpump",
  "tokenSymbol": "DIAMOND",
  "price": 0.000189,
  "price24h": 0.0000026,
  "mcap": 15.74,
  "featured": false,
  "scrapedAt": "2026-05-15T12:05:00.000Z"
}

---

🧰 Recipes

1. Live KOL alpha feed (60 min run)

{
  "feeds": ["twitter"],
  "userTags": ["kol", "master", "trader"],
  "twTypes": ["tweet", "quote"],
  "hasToken": true,
  "pollIterations": 60,
  "pollIntervalSec": 60
}

Polls every 60 seconds for 60 minutes. Skips duplicates. Webhook every new tweet to your alpha channel.

2. Exchange listing watcher

{
  "feeds": ["twitter"],
  "userTags": ["exchange", "binance_square", "exchange_listing"],
  "twTypes": ["tweet", "quote", "pin"]
}

3. Profile change / scam-detection feed

{
  "feeds": ["twitter"],
  "userTags": ["founder"],
  "twTypes": ["name", "handle", "description", "banner", "delete_post"]
}

Surfaces project founders renaming Twitter or deleting posts — strong scam signal.

4. Telegram alpha channels with high win-rate

Run with feeds: ["telegram"], then filter the dataset to channelWinRate > 0.5 and channelMembersCount > 1000.

5. Daily archive

Run with pollIterations: 1 once per day, schedule on Apify. Builds a long-term archive of KOL chatter for ML training.

---

🛡️ Under the hood

• Endpoints —
  • GET /vas/api/v1/twitter/messages?has_token=&user_tags=&tw_types=&featured=&mine=
  • GET /vas/api/v1/tg/messages?limit=30&featured=
• Polling — built-in iteration with deduplication via in-memory ID set.
• Auth — none (Cloudflare-fronted, fingerprint-gated).
• Proxy — Apify Residential default.
• Retries — 5 attempts with exponential backoff on 403 / 429 / 5xx.

---

⚡ Performance

Configuration	Rows	Time	Cost
Single poll, both feeds	50-150	~10s	<$0.01
30 polls × 30s (15 min)	200-1000	~15 min	~$0.03
60 polls × 60s (60 min)	500-3000	~60 min	~$0.08

---

📅 Recommended schedules

• Continuous live feed — pollIterations: 60, pollIntervalSec: 60 for hour-long runs scheduled every hour.
• Hourly snapshot — single poll every hour.
• Daily archive — single poll once per day.

---

❓ FAQ

How does GMGN classify KOLs?

GMGN curates a list of ~10k crypto Twitter accounts and tags them by user_tags. The classification is internal but stable.

Does this give me EVERY tweet from KOLs?

GMGN's feed is a cursor-less most-recent-N feed. Polling continuously catches everything new without misses. The feed depth is ~50-100 most-recent items per call.

Why are some tweets in Chinese?

GMGN was founded by Chinese-speaking devs; translation.zh-CN is provided alongside originals.

How is hasToken determined?

GMGN parses the tweet for cashtags and contract addresses, then resolves to a chain + token. Only matches set tokenAddress.

What's channelWinRate on Telegram?

GMGN tracks each channel's signal calls and computes a public win rate (0-1). Above 0.5 = better than coin-flip.

Can I filter by specific KOL usernames?

Not directly — but post-process the dataset by userScreenName. To monitor a specific user only, see Apify's Twitter Scraper actors.

How do I get a live webhook for every new tweet?

Use Apify's webhook integrations — each pushData triggers a webhook with the new rows.

Why are profile-change events (name, handle, ...) useful?

Scam projects often rename / delete posts to cover their tracks. Watching delete_post events on userTags: ["founder"] accounts is a strong scam signal.

Does it catch deleted tweets?

Yes — tw_type: "delete_post" is emitted with the original tweet_id reference.

Can I scrape historical KOL feeds?

GMGN only exposes current state. To build history, schedule the actor regularly and archive results.

What if a tweet has no sourceContent?

For pure profile-change events (banner / handle / name), sourceContent may be empty — the event itself is the signal.

---

🔗 Related actors

• GMGN Trending Memecoin Scanner — pair with this to find which tokens KOLs are talking about.
• GMGN Smart Money Buy Signals — correlate KOL tweets with on-chain whale activity.
• GMGN Token Holders Intelligence — see if KOLs themselves are holding the tokens they shill.
• GMGN New Token Launch Detector — fresh launches first heard about on Twitter.
• GMGN Wallet Portfolio & PnL Tracker — audit a KOL's actual PnL.
• GMGN Smart Money Wallet Leaderboard — top traders.
• GMGN Token Rugcheck & Deep Analytics — security audit per token.

---

🆘 Support

Bug? Want a new user-tag exposed? Open an issue on the actor's Apify page.