HostDeFi › Guides › Launch statistics
How many tokens launch on Solana per day? The measured numbers
We stopped quoting other people's estimates and started counting. These are our own capture monitor's dated figures — one ledger row per token, written the moment it launches.
Between 20,912 and 35,663 new tokens launched on Solana's launchpads every single day in the first week of September 2026 — 151,586 token launches in five days, roughly one every three seconds at the week's pace and one every 2.4 seconds on the busiest day. Most launch statistics you will see quoted are secondhand and undated; the counts on this page are primary measurements, each tied to the UTC day it was captured, so when a figure here is cited the date travels with it. And the distribution of who does the launching is the most useful thing in the whole dataset, so this page leads with it: those 151,586 launches came from just 37,910 distinct wallets, and 82% of the launches came from wallets that launched more than one token in the window.
The daily numbers
| UTC day (2026) | Token launches | Distinct wallets | From repeat wallets | From wallets launching 5+ | Busiest wallet |
|---|---|---|---|---|---|
| Sep 1 | 35,663 | 10,553 | 80% | 65% | 473 launches |
| Sep 2 | 33,227 | 10,514 | 77% | 63% | 923 launches |
| Sep 3 | 32,825 | 9,036 | 81% | 68% | 440 launches |
| Sep 4 | 28,959 | 8,559 | 79% | 65% | 449 launches |
| Sep 5 | 20,912 | 6,896 | 76% | 62% | 316 launches |
Every row reads the same way: tens of thousands of token launches, a much smaller pool of wallets doing them, and three-quarters or more of each day's output coming from wallets that launch again and again. September 5 was the quietest of the five days and still cleared twenty thousand. For scale context from outside our own data: industry reporting in mid-2026 described single days above 40,000 launches and put pump.fun's lifetime total near 11.9 million tokens — figures we cite as reported rather than measured, since our own ledger begins on August 31, 2026.
Where the launches happen
Of the 35,663 rows captured on September 1, 2026, our monitor stamped 23,984 with an explicit pump.fun bonding-curve origin and 79 with a bonk.fun origin; the remaining 11,600 arrived over the chain-level stream, which watches the pump.fun program directly and does not stamp a venue label. In practice, the overwhelming majority of Solana token creation runs through pump.fun-style bonding curves — which matters for safety reading, because those programs revoke mint and freeze authority at creation automatically. Our guide on mint and freeze authority covers when that flag is meaningful and when it is table stakes, and the pump.fun launch risk profile walks the whole birth environment.
The token factories
The launch counts only make sense once you see the cadence. The busiest wallet of the window created 923 tokens on September 2 alone — a median of 28 seconds between launches, sustained across the day. September 4's leader ran even hotter at a 21-second median gap. Over the full five days the top wallet shipped 1,462 tokens. This is industrial output: scripts minting tokens the way a spam botnet sends email, betting that one in a thousand catches a wave. Meanwhile 27,144 wallets — 72% of all launch wallets — launched exactly once, yet accounted for under a fifth of the launches. The launch feed you scroll is mostly factory product with a thin layer of genuine one-off attempts mixed in.
83% of launches share a ticker with another launch
Of the 151,586 token launches captured, 125,853 — 83% — used a ticker that at least one other launch also used inside the same five days. The ticker “SOL” was applied to 1,189 separate new tokens; “CAT” to 585; single trending words routinely spawn hundreds of same-ticker launches within hours of a news event. The practical consequence is blunt: on Solana a ticker identifies nothing. Only the mint address does. If you buy by ticker, the odds are good that you are choosing among dozens of unrelated tokens wearing the same three or four letters — which is exactly the mechanism behind the clone scams our token impersonation guide documents, and why verifying the contract address is step one of any check.
How we measure — and what we deliberately don't claim
The monitor listens to launchpad websocket streams and a chain-level subscription on the pump.fun program, de-duplicates by mint address, and appends one compact row per launch — mint, symbol, deployer wallet, venue, capture timestamp, market cap at capture — to a per-day ledger. A day's file closes at UTC midnight, and only closed days are reported here. Three honest limitations. First, the ledger counts what the monitor observes: a stream outage undercounts, and we would rather publish a dated undercount than an interpolated guess. Second, coverage began August 31, 2026, so lifetime totals are outside what we can measure. Third, a launch is not a listing — nearly all of these tokens will never reach a real market, which is exactly why per-launch structural checks exist.
You are welcome to cite these numbers. Quote any figure together with its UTC day, attribute it to HostDeFi launch-capture data, and link this page as the source; each figure stays pinned to its date rather than being silently replaced. The raw per-day exports behind this page are public and free through our API — if you want a cut we haven't published, compute it yourself.
Why the count is worth watching
Launch volume is the denominator of every meme-coin safety claim. When someone says "most new tokens are scams," the interesting question is: most of how many? At thirty-plus thousand launches a day, even a tiny fraction of predatory launches is a large absolute number, and no human can review them by hand — the argument for automated, per-token structural checks rather than vibes. Our graduation guide covers what the funnel from launch to real market actually looks like.
What a buyer does with these numbers
Treat every fresh launch as factory output until it proves otherwise, because statistically it is. The checks that separate the real attempt from the assembly line are the ones the factories don't bother with: holder distribution beyond the deployer, liquidity that is actually committed — see what liquidity locks prove and don't — and behavior in the first hour, which the first-hour checklist covers. The memecoin safety checklist strings the whole sequence together.
Check any launch against the ledger's lessons
Paste a mint address — the scan reads the on-chain state those thirty thousand daily launches are judged by.
Frequently asked
How many tokens are launched on Solana per day?
Across September 1–5, 2026, HostDeFi's capture monitor recorded between 20,912 and 35,663 new token launches per day on Solana's launchpads — 151,586 over the five days, roughly one every three seconds at the week's pace. Each figure on this page is pinned to its UTC capture day.
Who is launching all these tokens?
A small population of high-volume wallets. Over the same five days, 82% of launches came from wallets that launched more than one token, and 71% came from wallets that launched five or more. The single busiest wallet created 1,462 tokens in five days — at its peak, one every 28 seconds, median, sustained for a day.
How often do new tokens reuse an existing ticker?
Constantly: 83% of the 151,586 launches shared their ticker with at least one other launch in the same five days. “SOL” alone was used as a ticker on 1,189 separate new tokens. A ticker identifies nothing on Solana — only the mint address does.
How many pump.fun tokens ever graduate?
Industry reporting in mid-2026 put graduation — a token completing its bonding curve and moving to a real DEX market — at under 2% of all launches. Our own ledger is newer than that claim, so we cite it as reported rather than measured.
Where does this data come from, and can I cite it?
From HostDeFi's own capture monitor, which subscribes to Solana launchpad streams and the pump.fun program itself and writes one de-duplicated row per token the moment it launches — primary measurements, not aggregated estimates. Yes, you may cite any figure: quote it with its UTC date, attribute it to HostDeFi launch-capture data, and link this page. The daily raw exports are public and free.