How to Scrape Trending Memecoins Across Chains in 2026

New launches are a firehose and smart-money buys are a point signal — but the question most traders and dashboards actually start with is broader: what’s trending right now, across every chain, and is it real? Trending boards are where attention concentrates, and attention is what moves memecoin prices. The hard part is that “trending” is multi-dimensional: a token hot on the 1-minute board may be a momentary bot-pump, while one trending on the 24-hour board across multiple chains is a genuine market rotation. This guide covers how to scrape trending boards across six chains and four timeframes with enough per-token context to separate signal from noise, and how to wire it into bots, dashboards, and research.

What’s worth extracting

A trending board entry is only useful if it comes with the context to judge it. For each trending token, across each timeframe, you want:

• Identity — name, symbol, address, chain, and the timeframe it’s trending on.
• Pricing and momentum — current price plus price change across the relevant windows (1m, 5m, 1h, 24h). The shape of the move matters as much as its size.
• Market structure — market cap, liquidity, volume, and circulating supply. A 400% move on $3K liquidity is not the same as one on $300K.
• Holders — holder count and distribution, the base on which momentum either sustains or collapses.
• Smart-money exposure — smart-money and KOL count on the token, the single best filter for “is this attention validated or manufactured?”
• Manipulation signals — bot rate, sniper rate, bundler rate. High values mean the “trend” is a coordinated pump, not organic demand.
• Security — honeypot and other security flags, taxes, and dev/creator metrics, so you don’t chase a trending trap.
• Provenance — launchpad and exchange metadata, social links, duplication signals (is this a copycat of a hotter token?), and GMGN’s own scoring/rank data.

That’s 80+ structured attributes per token — dense enough to filter a trending board down to validated, tradeable candidates.

Why multi-timeframe and multi-chain together

Single-timeframe trending is a trap. The 1-minute board is dominated by transient bot pumps; the 24-hour board lags real-time rotation. The signal lives in agreement across timeframes — a token climbing the 5-minute and the 1-hour board is sustaining, not spiking. And memecoin attention rotates between chains: liquidity that leaves Solana shows up on Base or BSC days later. Scanning all six chains together is how you catch that rotation before it’s obvious.

This actor queries GMGN’s trending/rank boards across Solana, BSC, Ethereum, Base, Tron, and Monad and four timeframes (1m, 5m, 1h, 24h), optionally merging duplicates across timeframes, and can include a curated BSC rank. As with the rest of the suite, the data backs GMGN’s web front-end with no public API, so the actor uses a browser-like fingerprint and rotating residential proxy, bounded concurrency, and exponential-backoff retries to pull it reliably.

▶ Run the GMGN Trending Memecoin Scanner — trending tokens across six chains and four timeframes with 80+ fields each: price, volume, liquidity, holders, smart-money count, and security audit. Pick timeframes and security filters.

Schema design for downstream use

Normalize each trending entry into a row you can filter, rank, and alert on:

{
  "chain": "base",
  "timeframe": "1h",
  "token_address": "0x9f...c1",
  "symbol": "BRETT2",
  "rank": 3,
  "price_usd": 0.0182,
  "price_change_1h_pct": 64.0,
  "market_cap_usd": 1900000,
  "liquidity_usd": 240000,
  "volume_1h_usd": 510000,
  "holder_count": 4120,
  "smart_money_count": 9,
  "kol_count": 2,
  "bot_rate": 0.12,
  "sniper_rate": 0.08,
  "is_honeypot": false,
  "gmgn_score": 78,
  "scraped_at": "2026-06-02T15:00:00Z"
}

Schema choices worth making early:

• Always store timeframe and rank together. A token’s meaning depends entirely on which board it’s trending on — key your time-series on both.
• Keep smart_money_count first-class. It’s your cleanest filter: trending + smart-money-validated beats trending alone.
• Persist the manipulation rates (bot_rate, sniper_rate, bundler_rate). They’re how you reject a “trend” that’s really a coordinated pump.
• Store liquidity_usd alongside any price-change field. A huge move on thin liquidity is unexitable; the pair tells the real story.
• Log scraped_at. Trending boards turn over in minutes; a row is a snapshot, and your archive of snapshots is the trend.

Typical use cases

• Discovery and monitoring — a live cross-chain view of which memecoins are heating up, filtered to your liquidity and security thresholds.
• Trading bot / sniper feeds — pipe trending lists, gated on smart-money count and security, into automated entry logic.
• Top-N alerts — auto-post hourly or real-time “Top 10 Trending” rolls to Telegram/Discord channels.
• Multi-chain dashboards — track liquidity migration, market rotation, and launchpad performance across chains in one view.
• AI / quant datasets — build labeled training sets of on-chain profile-at-trend vs subsequent outcome for trend-classification models.
• Backfilled research archives — store daily trending snapshots for backtesting which trend signatures actually pay.
• Data resale and analytics — provide early-warning trend feeds to funds and analytics teams.

The common thread: value is in validated, multi-dimensional trending. A raw trending list is a casino board; trending cross-referenced with smart-money, liquidity, and security is an actionable shortlist.

Cost math for the managed approach

Pricing is pay-per-event: $0.00005 per actor start plus $0.0025 per result row.

Worked examples:

• An hourly “Top N trending” alert pulling the top 50 per chain across six chains, merged across timeframes (~300 rows/run): 24 runs/day is ~7,200 rows/day, about $18/day at full coverage — far less if you scope to two or three chains and one timeframe.
• A focused Solana-only 1h+24h scan, top 30 each (~60 rows/run), every 15 minutes during active hours: roughly 5,800 rows/day, about $14/day.
• A daily trending archive across all chains and timeframes for research (one larger run) lands in the low single digits per day.

The lever again is scoping to the chains, timeframes, and depth your strategy uses rather than pulling everything. A self-built equivalent means maintaining per-chain trending endpoint scrapers, the residential-proxy fingerprinting to survive them, and the smart-money/manipulation enrichment that makes the data worth anything.

Common pitfalls

• Single-timeframe tunnel vision. The 1m board is mostly bot noise; require agreement across timeframes for a real trend.
• Ignoring liquidity. A 300% move on $2K liquidity can’t be exited at size — always read price-change against liquidity_usd.
• Chasing unvalidated trends. Trending with zero smart-money and a high bot rate is a pump-and-dump in progress. Gate on smart_money_count and the manipulation rates.
• Missing the security check. Trending honeypots exist precisely because trending attracts buyers. Filter on security flags.
• Over-pulling all six chains when you trade one. Scope the run; you pay per row.
• Treating a snapshot as a trend. One pull is a moment; the value is in the time-series of snapshots showing momentum building or fading.

Wrapping up

Trending boards are where attention — and therefore price — concentrates, but raw trending is a casino floor until you cross-reference it with smart-money validation, liquidity, manipulation rates, and security. Building that across six chains and four timeframes means maintaining a fleet of proxied scrapers plus the enrichment layer. If you want validated, multi-chain, multi-timeframe trending data without owning any of that, use a managed actor that already normalizes GMGN’s trending boards.

▶ Open the GMGN Trending Memecoin Scanner on Apify — trending tokens across six chains and four timeframes, 80+ fields each, with smart-money and security context. Pay $0.0025 per token. Start with Apify’s free monthly credit.