Quick answer
In three random samples totalling 160 Solana launches created 120 to 30 days before Oct 8, 2026, exactly one token settled above $1M. That is about 0.6%, with a plausible range of roughly 0.1% to 3.5% because the sample is small. The median launch peaked at about $2,100 to $2,300 in market cap. Settling means at least two hourly closes, not a single candle high.
What is the real base rate for a memecoin reaching $1M?
Close to 1 in 160 in our samples, and probably lower than most traders assume. The headline figures:
The one token that made it, OUSD, crossed on July 2, peaked at about $2.28M and later fell to zero. So even the lone success round-tripped below the line.
Data, not financial advice. This is a base rate, a description of how rare the outcome is, not a forecast for any token.
What did the three runs show?
Each run used stalkchain_milestone_study with the launches cohort. That cohort is every token created in the window, including the dead ones, which is the only frame that gives an honest denominator.
| Run | Threshold | Tokens studied | Crossed | Median lifetime peak |
|---|---|---|---|---|
| A (earlier readout) | $1M | 40 | 0 | $2,294 |
| B | $1M | 60 | 1 | $2,127 |
| C | $1M | 60 | 0 | $2,119 |
| Pooled | $1M | 160 | 1 | about $2.1-2.3k |
| D | $100k | 58 | 1 | $2,251 |
The pooled rate is 1 in 160, or 0.63%. A 95% Wilson interval on one success in 160 runs from about 0.1% to 3.5%. Put plainly, the true rate could be a tenth of what we measured or several times higher. We cannot say more than that.
An earlier 8-token run at a $3M threshold found zero crossers and, with so few tokens, could say nothing else. That is the reason we raised the sample size.
How was it measured?
The call and its parameters, as you would type them into Claude or ChatGPT:
Of Solana tokens launched 120 to 30 days ago, how many ever
settled above $1M market cap? Sample 60, launches cohort.
stalkchain_milestone_study
thresholdUsd: 1000000
cohort: "launches"
sample: 60
daysAgoFrom: 120
daysAgoTo: 30
Pulled Oct 8, 2026, about 07:15 UTC via StalkChain's MCP (stalkchain_milestone_study). Trimmed response for run B:
{
"thresholdUsd": 1000000,
"cohort": "launches",
"window": { "fromDaysAgo": 120, "toDaysAgo": 30 },
"cohortSize": 2950578,
"studied": 60,
"crossedThreshold": 1,
"crossedPct": 1.7,
"medianLifetimePeakUsd": 2127,
"wentHigherPct": 100,
"roundTrippedBelowPct": 100,
"peakMultiple": { "median": 2.28, "atLeast5x": 0 },
"tokens": [
{ "symbol": "OUSD", "crossedAt": "2026-07-02", "peakAfterUsd": 2279947, "endedUsd": 0, "roundTripped": true }
]
}
Same call over REST:
curl -G https://data.stalkchain.com/api/v1/tools/stalkchain_milestone_study \
-H "Authorization: Bearer sc_YOUR_API_KEY" \
-d thresholdUsd=1000000 \
-d cohort=launches \
-d sample=60 \
-d daysAgoFrom=120 \
-d daysAgoTo=30
Field by field:
cohortSize: how many launches existed in the window. It differed between runs (2.62M, 2.95M, 3.34M), so the index is still filling. Quote the sample rate, not the cohort count.crossedPct(1.7): one token in a sample of 60. It is a single observation, and the tool says so in its own notes. Do not quote 1.7% as the base rate.medianLifetimePeakUsd: the middle token's highest market cap ever. About $2,100 means half of all launches never got past the price of a used phone.wentHigherPctandroundTrippedBelowPct(both 100): these rest on one token. Ignore them at this sample size.
Why hourly closes and not candle highs?
Because candle highs lie on illiquid tokens. A token with a few thousand dollars of liquidity can print one trade that values the entire supply at a huge number. Read the highs and dead tokens look like thousand-fold winners.
So a token counts as reaching a line only if at least two hourly closes settled above it. One stray print does not count. That rule is why $1M here means a market cap that held, however briefly, rather than a wick.
It also means our rate is conservative on purpose. Some tokens touched $1M for a minute and are not counted. That is the point.
What does the "reached" cohort change?
The tool offers a second cohort called reached. It draws from tokens that got near the milestone, so it finds about seven times more crossers per call. The catch is survivor bias. It over-represents tokens that stayed high, so any "went higher" share from it is an upper bound.
Use launches to answer "how rare is $1M?". Use reached only to study what winners did after they crossed, and remember the winners you see were selected for staying visible.
We used launches for every number on this page.
Why is n=160 still small?
Because the event is rare, and rare events need large samples. With one success in 160, a single extra or missing success moves the rate a lot: 0 of 160 would be 0%, and 2 of 160 would be 1.3%.
We also ran a $100k threshold on 58 tokens (two of 60 had no chart data and were dropped). One token crossed, USTF, which peaked near $1.6M on a 16x move from the line and later ended around $1,248.
A $1.6M peak means it would likely have qualified at $1M too. These outcomes are lumpy: one token can decide a whole cell of the table.
Other limits:
- Random samples vary. Rerun it and you will get a different draw.
- One window. Tokens created 120 to 30 days ago, on Solana only.
- Overlap is possible. Separate random draws could include the same token, though the odds are low in a pool this large.
- Market cap basis. The study measures market cap from price data, not liquidity you could actually sell into.
To measure what happens after a token crosses, you would need roughly 1,000 or more samples, which is about 17 calls at the maximum of 60. That is possible, and a monthly refresh is a sensible way to build it.
What does the base rate mean for you?
Use it to calibrate expectations, not to pick tokens.
If a launch reaching $1M is roughly a 0.1% to 3.5% event, then a token pitched as "the next $1M" is asking you to believe in a rare outcome. Rare outcomes happen, but they are not the default, and the median launch topped out near $2,000.
An illustration with heavy uncertainty: at 0.1% of a 3.3M-launch cohort you would still get about 3,300 tokens, and at 3.5% about 117,000. The point is that even a tiny rate across millions of launches produces visible winners, which is why survivorship makes the game look easier than it is.
Practical habits that follow:
- Size positions assuming the typical outcome is near zero.
- Check the basics before the chart: the deployer, snipers and bundles, and the due diligence checklist.
- Treat a milestone as a fact about the past, not a signal about the future. In our one example the token that crossed went to zero.
How do you run this yourself?
Connect the StalkChain MCP server to Claude, ask the question above and adjust the threshold, window or sample. The Claude page for Solana data is at /ai/claude/solana, and the data set is described on Solana data for AI agents.
A study costs one upstream call per token studied, so start small. New accounts get $5 of free credit with no card.
Methodology
- Tool:
stalkchain_milestone_study, StalkChain MCP, pulled Oct 8, 2026. - Cohort: every Solana token created 120 to 30 days ago (
launches), dead ones included. Cohort sizes 2,619,644, 2,950,578 and 3,343,953. - Sample: random, 40 + 60 + 60 = 160 tokens at $1M; 58 tokens at $100k.
- Crossing rule: at least two hourly closes above the threshold. Candle highs are not used.
- Statistic: pooled crossing rate, 95% Wilson interval, median lifetime peak per run.
- Not claimed: anything about what happens after a token crosses. The sample has one observation.
- Refresh: we plan to rerun this monthly with larger samples.
FAQ
What percentage of memecoins reach $1M market cap?
In our pooled sample, 1 of 160 random Solana launches (0.6%) settled above $1M, with a plausible range of about 0.1% to 3.5%. "Settled" means at least two hourly closes above the line. The sample is small, so treat this as an order of magnitude rather than a precise rate.
What is the median peak market cap of a new memecoin?
About $2,100 to $2,300 across our three $1M runs. Half of all launches in the window never exceeded that. It is the lifetime high, not the current value.
Why use hourly closes instead of the highest price?
A single candle high on a thin token can value the whole supply at a fake number. Requiring two hourly closes filters out those prints, so dead tokens are not counted as 1000x winners.
Is the "reached" cohort a good base rate?
No. It samples tokens that already got near the milestone, so it is biased toward winners and its "went higher" share is an upper bound. Use the launches cohort for a base rate.
Can you tell if a token that hit $1M will go higher?
Not from this data. Our one crosser peaked at about $2.28M and then went to zero, and a single observation proves nothing. Measuring it properly needs roughly 1,000 or more sampled launches.