Copypasta

TL;DR

A hot ticker on Solana is not one coin. It is dozens of coins with the same name, launched minutes apart by bots and copy desks. On 2026-10-09, pump.fun's search held at least 61 Solana coins with the ticker $FOMO (a reviewer's deeper read of five search pages found 92). 26 of them ran past $50K, and 24 of those 26 are now down 85% or more. Only one held real liquidity: a coin launched two days earlier, sitting behind a swarm of 54 same-ticker launches that week. Traders who buy "$FOMO" off a tweet, a chat or a trending list usually can't tell which contract that is.

Copypasta answers the question traders ask in that moment: "which $TICKER is the real one, and what happened to the copies?" Type a ticker or paste a contract. You get every same-ticker coin we can find and three reference coins: the leader (most real liquidity right now, with fake caps and parked pools removed), the first to run this week and the first ever. Every other coin gets a clone number and an outcome: rugged, instant self-bond, alive, on the curve or dead. A worker logs every pump.fun launch each minute. It powers a live board of the tickers being copied right now, a "fresh copies" feed, an hourly copy rate and a serial-cloner leaderboard.

Free tools exist (OGfinder, GMGN's "same name" panel). They sort lookalikes by age. Our data shows age is the wrong answer: the oldest $FOMO is a dead December-2024 Raydium coin, and the oldest $NASA isn't the leader either. Copypasta is built around where the money is now, what each copy did (clone tax, instant bonds) and who mass-produces copies.

Token: $PASTA (reserved, not launched). The token is the brand of a meme-able product. Creator fees fund the data bill, holders get alerts later (read-only balance check), and Copypasta will show its own clone report for $PASTA.

The user and the pain

Who: the pump.fun / PumpSwap trader who buys narratives fast, at $20K–$5M market caps. They act off an X post ("$FOMO is sending"), a Telegram call, a stream, or a trending list on Axiom, GMGN, Photon or a bot. The ticker reaches them before the contract address does.

The pain:

• The ticker is ambiguous by design. When a name trends, launch bots and copy desks deploy the same ticker within minutes. Our launch log caught one wallet deploying $QSB 9 times, $Muskrat 4 times and 6 other hot tickers within about 20 minutes. Another created 8 "Fomo Cat" coins in two hours.
• Copies borrow legitimacy. Of the 61 $FOMO coins, 28 link to the same X profile. For $TikTok, 37 coins link to the same tweet. Copies look official because they point at the same source.
• Copies pump, then die. Many copies are instant self-bonds: the creator buys out the curve, the chart prints a fast "ATH" (FOMO copies show $400K–$3.8M ATHs), and the bag is dumped on whoever arrives. Those coins now sit at about $2K.
• "Oldest" is not "real". The usual heuristic is "the oldest one is the OG". It fails when an old dead coin shares the ticker, or when a newer coin is where the community actually trades.
• Afterwards, people want proof. "I bought clone #14" or "the real one is this CA" is a post people make every day. Without a tool, they screenshot a DexScreener search.

The moment of use is short: a ticker is moving and they have seconds. The return loop is the board (what's being copied right now), the watchlist (what changed on my tickers since last time) and the cloners page (who's doing it).

Evidence

• Copycats are a measured, massive share of pump.fun. "Meme Coin Factories" (arXiv 2609.10246, 2026-09-09) analysed 15,245,966 pump.fun coins from Jan 2024 to Jan 2026. On the strictest definition (same name, ticker, description and image), 1,490,123 coins (~10%) copy 647,986 originals. Matching name and ticker only gives 5,392,081 (36%, with some false positives). Copycats graduated 0.86% of the time vs 9.20% for originals. The top 1% of wallet clusters made 53–59% of all coins, and the median post on X or Truth Social spawned 12 coins. (arXiv)
• Same-ticker runs split real money across contracts. In the $ANSEM episode (late June 2026), several $ANSEM tokens from different anonymous creators ran at once and traders flipped between them; the leader peaked around $60M. (crypto.news, 2026-06-29)
• Name-squatting hits outside memes too. At least 9 tokens used St. Jude's name and logo within 72 hours of pump.fun's charity-coin launch. The article notes there is "no pre-deployment consent check". (TechBuzz, 2026-05-04)
• Copies crash hard. A digest of the October Zcash-meme wave reports that "several copies are already down 58% to 89%" (the copycat launches that followed ZCAT, which itself peaked near $170M). (U.Today, 2026-10-07)
• The copycat pattern is standard scam guidance. Guides describe "REALTOKEN2 / RealToken Official" copies appearing within minutes of a trend, and advise checking the CA against official channels. (ChainLens, date unknown; Webopedia, date unknown)
• Traders blame copycat "PvP" for the weak meme market. (SolanaFloor, about 5–6 months old)
• Historic scale (older): in January 2025, Cointelegraph (citing Birdeye) counted 61 fake TRUMP/MELANIA tokens that took in $4.8M from 12,641 wallets in 24 hours. (TradingView/Cointelegraph, Jan 2025)
• Losses are the base rate. Solidus Labs found 98.6% of pump.fun tokens (Jan 2024–Mar 2025) ended as rugs or pump-and-dumps. (Solidus Labs, 2025)
• Not found / unverified: a dated figure for total trader losses from buying the wrong same-ticker contract. A Financial Times count of 167 Trump copycats appeared only in a search summary and is not used here.

Market now

Live numbers, all dated 2026-10-09:

• Launches: the factory's pump.fun digest (14:19 UTC) estimates ~54,900 launches/day and 1,050 graduations of coins created in the last 24h, a bond rate of about 1.9%. 317 of the 1,050 were instant self-bonds (ATH within 2 minutes of creation, then −50%). Copypasta's own log captured 718 launches between 14:34 and about 14:55 UTC, roughly 31 per minute (~45K/day at that pace).
• Clone clusters among graduates (digest, 24h): FOMO ×34, Google Gemini ×19, MrBeast ×12, Anthropic ×12, Tiktok Coin ×10, claude ×9, ChatGPT ×8, NVIDIA ×7. Most of the "ran then went to zero" list is name and brand copies: three separate $FOMO coins at $3.5–4.4M ATH are now at ~$2K.
• Copypasta cluster reads (about 14:40–14:55 UTC):

• $FOMO: leader kZXne…pump with $242K liquidity; 28 coins link x.com/fomo; 1 instant bond.

• $NASA: the leader was created at 13:00 UTC that day; 50 coins link x.com/nasa; 3 instant bonds.

• $TikTok: 37 coins link the same tweet; 8 instant bonds. $PQC: one creator launched it 7 times; 10 instant bonds. $Gary: 15 wallets launched it 2+ times.

• The live board's first snapshot (14:42 UTC): 11 hot tickers. 18 of their coins ran and 11 of those rugged, a clone tax of 61%.

• Copy rate: in the first ~20 minutes of logging, 151 of 490 launches (31%) reused a ticker that 3+ coins from 2+ wallets had already used. For the whole 14:34–14:55 UTC window it was 329 of 718 (46%). That figure uses a lookback shorter than 24h, so it is not yet a steady-state rate. One wallet made 27 copies across 6 hot tickers in that window.

Competitors and the gap

The gap: nobody answers "which one holds the money right now" with fake caps and parked pools filtered out, and nobody shows what happened to the copies as a number (the clone tax). Nobody runs a public, live board of tickers being copied right now, or a leaderboard of the wallets producing copies across all tickers. Copypasta also gives a shareable verdict ("CLONE #9 of 54") with a card.

The product

Live at /p/copypasta/app/. Hash routes; share links use ?t=<ticker> and ?ca=<mint>.

Views

1. Board (home, before any input): live stats (launches logged, copies on the board, clone tax, copies in the last 10 minutes). Cards for the 16 most-copied tickers of the last 24h, each with launches per 15 minutes over the last 6h, coins found in 7 days, how many ran, rugged, were instant bonds and are still trading, the leader with its liquidity, and a clone-tax bar. A "+n new" badge shows launches since the snapshot. Also a Fresh copies feed (refreshes every 20s) and an hourly copy rate chart.
2. Ticker page (#/t/fomo): a headline sentence, the three reference cards (Leader, First to run or Biggest fall, First ever), a swarm chart (each coin plotted as launch time × all-time high, coloured by outcome), counts, launches per hour from our own log, wallets that launched the ticker more than once, and shared X links. Below that, the full table with filters (all, alive, ran, rugged, dead). Watch button, share card and copy link. It re-pulls every 90s.
3. Coin check (#/ca/<mint>): an animated verdict stamp (THE LEADER / FIRST TO RUN / FIRST EVER / CLONE #n) with plain-English reasons. These cover how far it is behind the leader in liquidity, how long after the first runner it launched, its drawdown, instant bond, serial creator, fake cap or idle pool. A side-by-side panel compares it with the leader, plus a share card PNG and "copy leader's CA".
4. Serial cloners (#/cloners): wallets with the most copies of hot tickers in the last 24h of the log, which tickers they copied, launch counts and outcomes. Wallets stay addresses; nothing links them to people.
5. Watchlist (#/watch): tickers saved in localStorage, re-checked live, with a since-last-visit diff (new copies, rugs, leader changed), plus recent checks.
6. How: every definition and threshold, in plain words.

How it works

• Grouping: NFKC-normalized lowercase letters and digits of the ticker ("$FOMO", "F.O.M.O" → fomo); a coin whose full name equals the key also joins.
• Sources: pump.fun coins/search-v2 (3 pages of 50 newest-first plus the top 50 by market cap; EVM results dropped), DexScreener search (non-pump coins), DexScreener tokens/v1 for every graduated coin's best pair, pump.fun coins-v2/{mint} for a checked coin, and Helius DAS as a last fallback.
• Liquidity: for graduated coins, twice the backing side (SOL/USDC) of each DexScreener pool, summed over pools (Jupiter's aggregate for verified mints); for curve coins, the curve's real SOL × SOL/USD (pump.fun sol-price).
• Sanity filters: a cap more than 150× its pool (and over $250K) is a fake cap: excluded from "now", the ATH is marked "fake?", and the coin can't lead. A pool with $50K+ liquidity and under $1K traded in 24h is an idle pool and can't lead.
• Outcomes: ran = ATH ≥ $50K. Rugged = ran and now < 15% of ATH. Instant bond = graduated with the ATH within 120s of creation. Alive = graduated, ≥ $5K liquidity, ≥ $15K cap, ≥ $1K 24h volume. On curve = traded in 6h and worth ≥ $4K. Just launched = on curve, under 2h old.
• Backend (worker + SQLite): launches (every 60s) logs the newest 50–100 pump.fun launches (key, symbol, creator, time, X link, art) for 72h. board analyzes the 16 most-copied tickers (4+ launches from 2+ wallets in 24h) every 15 minutes and stores snapshots for 14 days. hourly rolls up the copy rate. Routes: /board, /fresh, /velocity?key=, /cloners, /history, /stats. Px use is about 120/h for launches and about 320/h for the board, inside the 1,000/h budget.

Calibration (hand-checked on real inputs, 2026-10-09):

• $FOMO: the engine picked kZXne…pump as leader. pump.fun's coins-v2 reports canonical-pool liquidity of $241,739 and the X link x.com/fomo; DexScreener shows a $241.7K pool with $189K of 24h volume. The 9 graduated copies created 09:09–12:22 UTC show ATHs of $406K–$3.82M and caps of ~$2K, all classed rugged. The "first ever" $FOMO (ZxBon…, Raydium, Dec 2024) correctly isn't the leader.
• $CALI: the first version picked 5V6kN…, an Orca pool showing $1.96M liquidity with $1.05 of 24h volume. That's parked liquidity, so the idle-pool rule was added. The leader is now 8k4sB…, a PumpSwap pool with $169K liquidity, $465K of 24h volume and 2,308 buys / 1,939 sells (DexScreener, checked by hand).
• BONK (DezXA…B263): first came back with no leader. pump.fun indexes BONK as "not complete" with an empty curve, and its $292M cap is over 150× its best single pool. Two fixes followed: an empty curve plus real DEX pools now counts as graduated, and coins 30+ days old with $100K+ of 24h volume are exempt from the fake-cap rule. Liquidity is now summed across all pools. BONK is now THE LEADER of 50 $Bonk coins.
• New curve coins minutes old were first labelled "dead". They're now "just launched" for 2 hours.
• $NASA / $TikTok / $PQC / $Gary: results in the Market table; leaders match the highest-liquidity actively traded pool on DexScreener for each.
• Coin check: 9EB7E…pump ("Fomo Coin") → CLONE #9+ of 52+ this week, ATH $78.8K, now $398 (−99.5%), launched 40h after the first $FOMO that ran.
• Reviewer fixes (2026-10-09 ~15:40 UTC), each checked against raw on-chain pool reserves:
  • Self-priced pools. Pasting the real USDC mint first named a fake "ERC20-USDC" Raydium CLMM pool as the leader: it showed $48.9M of liquidity backed by $21 of real USDC. Liquidity now counts twice the backing side (the SOL or USDC in the pool), and Jupiter-verified mints for the ticker rank first. USDC is now THE LEADER, and both fake "USDC" pools read about $42.
  • Dead-pool prices. DexScreener and pump.fun price a dumped coin at its last trade. 9EB7E…pump read $2.5K, but its PumpSwap pool (3.25 SOL against 862.6M tokens) prices it at $396. PumpSwap caps now use the reserve price when it is lower.
  • Trading coins below 15% of ATH. Gary the Cat (8ZCmw…pump) is −91% from a $4.62M ATH but traded $6.5M in 24h, and it was excluded from leader. As a result $Gary said "none holds real liquidity". Coins like this can now lead, and the leader card says they are down 85%+.
  • Silent truncation. Search pages hold 49 coins after filtering, so the "is the result capped?" check (all pages exactly 50) never fired. $FOMO showed 60 with no "+", while 92+ exist. Totals and clone numbers are now marked as floors ("59+", "CLONE #9+").
  • On-chain check after the fixes: $FOMO leader pool holds 1,089.6 SOL, which gives $238.4K liquidity (engine: $238.4K). $Gary leader pool holds 275.9 SOL, which gives $60.4K (engine: $60.4K).

Limits: pump.fun search caps results, so totals for tickers with hundreds of copies are a floor (marked "+"). Market caps and ATHs are third-party numbers with sanity filters on top. Grouping is by exact ticker ("$FOMOCAT" is its own group). The launch log starts at deploy: until it is 24h old, the board's 24h counts and the copy rate cover a shorter window, and the UI says so. pump.fun only; letsbonk/Bags launches appear only when DexScreener lists them.

Token design

$PASTA: reserved, not launched. No token exists unless listed at tekmaker.xyz/p/copypasta.

• Creator fees pay the data bill: the launch log, pump.fun and DexScreener reads, and, on the roadmap, an indexer for full search depth. This will be stated publicly on the project page.
• Holder tier (planned, read-only balance check, no signing): watch alerts when a new copy launches or the leader changes, deeper search (all pages), and cloner alerts. All of it is read-only.
• "Copy us, we'll catch you": $PASTA is a ticker made to be copied. Copypasta will show its own live cluster for $PASTA, with the real mint as the leader and every copy numbered. That turns the self-referential meme into proof the product works.
• No staking, no promised yield, no airdrop promises.

Launch plan

1. Seed the board on CT during a live clone wave: "61 coins share $FOMO. 24 of the 26 that ran are −85%. Here's the one with the money." Post the cluster share card.
2. Reply-guy distribution: when a ticker trends, reply with the Copypasta link (?t=). The card's "CLONE #n" stamp is the hook.
3. Publish a daily "clone tax" post from the board history, plus the top serial cloner wallets.
4. Launch $PASTA on pump.fun from one disclosed wallet with no bundle, then show its own cluster live (copies included).

Roadmap

• Telegram/X alerts for watchlist tickers (holder tier).
• Image-hash matching (same art, different ticker) when an image-hash source is available.
• letsbonk.fun and Bags launch logs.
• Leader-flip alerts: leader history is already kept in SQLite and shown per ticker; next is pushing a flip to watchers.
• Embeddable "real CA" badge for communities to post.

Risks and kill criteria

• Data access: if pump.fun search-v2 changes or rate-limits, fall back to DexScreener search plus our own launch log. Kill the project if neither works.
• Free competitor parity: if OGfinder or GMGN add liquidity-weighted leaders and outcome tracking, our edge shrinks to the board and cloner history. Kill or merge if, 30 days after launch, under 300 weekly users come back to the board.
• Wrong leader calls: third-party numbers can mislead (as the idle Orca pool did). Each new failure mode becomes a sanity rule, and the UI always says "leader ≠ safe".
• Harm: the cloner page shows on-chain wallets only, never identities. Nothing helps anyone launch copies.

Sources

• arXiv 2609.10246, "Meme Coin Factories" (published 2026-09-09, via research agent, accessed 2026-10-09)
• crypto.news, $ANSEM explainer (2026-06-29, accessed 2026-10-09)
• TechBuzz, charity coins (2026-05-04, accessed 2026-10-09)
• U.Today, meme digest (2026-10-07, accessed 2026-10-09)
• ChainLens, pump.fun scams guide (date unknown, accessed 2026-10-09)
• Webopedia, pump.fun scam tactics (date unknown, accessed 2026-10-09)
• SolanaFloor, PvP culture (about 2026-04, accessed 2026-10-09)
• TradingView/Cointelegraph, fake TRUMP/MELANIA tokens (Jan 2025, accessed 2026-10-09)
• Solidus Labs, Solana rug pulls report (2025, accessed 2026-10-09)
• OGfinder (fetched 2026-10-09)
• Jupiter token page docs (date unknown, accessed 2026-10-09)
• Solana Compass, RugCheck listing (date unknown, accessed 2026-10-09)
• GMGN blog, same-name tokens (about 2026-09, search summary only, page returned 403)
• Factory memeradar digest and deep pass, digest-latest.md and deep-latest.md (generated 2026-10-09 14:19–14:20 UTC)
• Live: pump.fun frontend-api-v3 (coins/search-v2, coins-v2, coins, sol-price) and DexScreener (tokens/v1, latest/dex/search) via the factory proxy, 2026-10-09 14:30–15:00 UTC