Decentralised exchanges list new tokens first, within minutes, because no review is required.
Among the eight centralised venues in this index, MEXC was the only one we found publishing how many tokens it lists each month, and the only one meeting all five disclosure tests.
It is also the quickest to remove a listing once it turns, which is the trade-off this report measures.
Key Takeaways
MEXC was the largest listing venue on centralised exchanges in the second quarter of 2026, accounting for 41% of all listings and more tokens than the next three exchanges combined, according to CryptoRank.
MEXC was also the only platform we found publishing its own monthly listing count, reporting 135 new tokens in March 2026 at 65% platform-first.
Listing speed itself cannot be verified from outside an exchange, so this index measures what can be checked: what each platform publishes about listings and removals.
Notice before spot trading stopped ranged from 1 day to 14 days across the eight platforms, and MEXC sits at the short end.
Withdrawal windows after a trading halt ran from 28 days at KuCoin and 30 days at MEXC to 92 days at Bitget, with Gate's notices setting no cut-off.
Only around 32% of newly listed tokens across the top 12 exchanges rise immediately after listing, per CoinGecko, so disclosure quality says nothing about whether a token is worth owning.
Every exchange advertises how many assets it supports.
Almost none will tell you how many it added last month.
That distinction decides whether a catalogue figure means anything, because a platform listing 3,000 assets could be adding 130 a month or three a month, and the catalogue number reads identically in both cases.
All eight platforms clear the first two tests, which is worth saying plainly.
Each runs a public, dated archive of new listings and a public, dated archive of removals, readable without an account.
The separation happens on the last three columns.
Seven of the eight publish a named framework for removing a token, with OKX the exception at the time of checking.
Five state what happens to a balance still sitting in the account after the withdrawal cut-off, and three do not.
One publishes a periodic count of what it actually listed.
We publish a monthly listing count because the number exchanges usually promote answers the wrong question.
A trader deciding where to hunt this week does not need to know how many assets a platform has accumulated since launch.
Independently, CryptoRank recorded MEXC as the largest listing venue in the second quarter of 2026, accounting for 41% of all centralised exchange listings and more tokens than the next three exchanges combined.
Those figures are self-reported, and that is worth saying directly, but we found no comparable periodic figure published by the other seven.
Run it forward across a quarter and the gap stops being cosmetic: a pipeline at 130 a month produces roughly 400 candidates in three months, one at three a month produces about nine, and both platforms can honestly print the same catalogue total on their homepage.
Now the part that does not flatter us.
MEXC and KuCoin run the two shortest withdrawal windows in this index, at 30 and 28 days.
Our published rule is that a token flagged under the ST warning is delisted three days later if the risk is judged severe, and withdrawals then stay open for 30 days.
Bitget, Bybit, OKX and Kraken all run roughly three months on the withdrawal leg, and Gate sets no cut-off at all.
We think a short cycle is defensible for the case it was built for, which is a project that has deteriorated to the point where leaving the pair open is itself a risk, and we would rather move quickly and publish the tag than let a failing market run quietly.
We also built one backstop that the shorter window makes necessary: if the 30-day withdrawal period is cut short because the project stops maintaining its chain, we convert the holding to USDT at the delisting price and notify affected users within 30 days.
What we are not going to argue is that 30 days is more generous than 92.
If you hold small positions you check once a quarter, a platform with a three-month window fits that habit better, and this report says so.
A delisting is not one event.
It is a sequence, and two intervals in that sequence decide what happens to your position.
This interval decides whether you find out in time to sell into a market that still exists.
The spread is wider than the platforms' similar-sounding language suggests.
Bitget's notices follow the most regular pattern in this group, with seven separate notices between March and June 2026 all published six or seven days ahead of the delisting date.
Bybit ran the same seven-day interval across the samples we checked, announcing the removal of ELON and VINU on 3 June 2026 for a trading halt on 10 June at 08:00 UTC.
Kraken's July 2026 cycle notice ran thirteen days, and the longest interval we recorded, at fourteen days, came from two Binance batches announced in June and August 2026.
MEXC sits at the short end, and the detail matters.
The 1-day figure comes from a trading-pair removal, where the pair closed but the token stayed tradable against other quote currencies, so the practical stakes are lower than a full token removal.
For full removals the published rule is three days from the ST tag, still the shortest published rule in this index.
Several platforms here label higher-risk tokens, and MEXC's ST tag is the one collected into a dated monthly table, so those three days do not arrive without context.
Bitget's withdrawal deadline lands exactly three months after each delisting date, without exception across the seven notices we checked.
OKX runs a similar three-month window measured from the announcement itself, and once the delisting completes the balance appears in the account under untradable assets.
Gate is the outlier in the generous direction.
Its notices set no withdrawal cut-off and state that the account continues to work as a wallet for the removed token, and it runs a separate buyback application window where holders can apply to have delisted tokens repurchased at stated prices, subject to a per-person cap.
The platform that delists the most also gives holders the longest exit.
At the other end, KuCoin closed withdrawals 28 days after a February 2026 delisting and 32 days after a June 2026 one, and MEXC's published rule is 30 days.
Kraken's model runs differently again, closing withdrawals after roughly three months and then liquidating whatever is left, with an explicit warning in its own notices that proceeds may be significantly below reference prices and in some cases minimal or nil.
Adding the two intervals gives the full window between the moment a removal is announced and the moment the balance can no longer leave.
The range runs from 33 days to 101 days.
For a trader who checks balances weekly, the difference is invisible.
For someone holding a small position they stopped watching months ago, which describes most airdrop and new-listing residue, it is the difference between recovering the asset and not.
Aggregates hide the shape of the thing.
Laid out as sequences, the same notices show how differently the stages are spaced: MEXC compresses everything into the first five weeks, Bitget and OKX hold a long quiet tail, and Kraken adds a fourth stage after the withdrawal deadline in which the platform itself sells.
That quiet tail is where positions get forgotten, which is why the withdrawal date matters more than the trading date for anyone not watching daily.
Platform | Dated new-listing archive | Dated delisting archive | Shortest verified notice before trading halt | Withdrawal window after halt | Periodic listing count published | Named removal framework |
MEXC | Yes, public | Yes, dedicated Delistings section | 1 day (trading pair); 3 days under the ST rule | 30 days | Yes: 135 listings in March 2026, 399 in Q1 2026 | Yes: Innovation Zone, Assessment Zone, ST warning |
Binance | Yes, public | Yes, public | 14 days | 61 days | Not published | Yes: Seed Tag and Monitoring Tag |
OKX | Yes, public | Yes, dedicated delistings section with publication dates | 4 days | About 85 days | Not published | None found on the official help centre |
Bybit | Yes, public | Yes, dedicated Delistings category | 7 days | About 91 days | Not published | Yes: Token Management Rules, Innovation Zone |
Bitget | Yes, public | Yes, in the support centre | 6 days | 92 days | Not published | Yes: Innovation Zone, periodic review |
Gate | Yes, public | Yes, public | 5 days | No cut-off stated, plus a buyback application window | Not published | Named review, plus a published buyback mechanism |
KuCoin | Yes, public | Yes, plus a standing table of closing times | Not verified | 28 to 32 days | Not published | Yes: Special Treatment Rules |
Kraken | Yes, public | Yes, monthly cycle notices | 13 days | 88 days, then automatic liquidation | Not published | Yes: scheduled delisting cycles |
Data verified as of 28 August 2026 against each platform's own dated public notices, except the Binance intervals, which are taken from contemporaneous media reporting because the underlying announcement pages could not be retrieved. Notice periods are the shortest interval observed in dated notices published between January and August 2026, not a guaranteed minimum. "Not verified" means the announcement date could not be established from the platform's own dated pages at the time of checking, not that no notice was given.
The literal answer is a decentralised exchange.
A project can open a liquidity pool without asking permission, so tokens trade on-chain minutes after launch, with no review, no support desk, and no recourse if the contract is a copy of the real one.
Among centralised venues the question splits into two cases that behave very differently.
For a high-profile token generation event, the major venues open within hours of one another.
When RE Token launched on 18 June 2026, four venues opened trading inside two hours, with MEXC and Binance Alpha at 12:00 UTC and KuCoin and Binance spot at 14:00 UTC.
Ranking those venues by hours is not useful information for a trader.
For the long tail of smaller tokens, the difference is not who lists first but who lists at all, which is a coverage question rather than a speed question.
So the version worth asking is narrower: when a platform lists something, does it tell you clearly, and does it tell you just as clearly when it takes the listing away?
The original design was a timestamp index: pick a neutral basket of tokens that reached their first centralised listing in one quarter, record each platform's listing time for each token, and report the median lag behind whichever venue was first.
Three things broke it.
Building a neutral basket requires a queryable record of tokens whose first centralised listing fell inside a given quarter.
Public listing trackers show rolling 30-day windows rather than historical quarters, and the analytics platforms holding the full database restrict automated access to their per-exchange listing pages.
Assembling a basket by hand from news coverage would bias the sample toward tokens that got press, which is exactly the bias that would distort the result.
Even with a basket, the median lag collapses.
For tokens the large venues do list, everyone lists on the same day, so the lag is near zero across the board.
For tokens the large venues never list, the lag is undefined rather than large.
A metric returning zero or undefined for almost every observation is not measuring anything.
Every platform in this index publishes dated announcements about what it lists and what it removes.
Those announcements carry timestamps, they sit on public pages, and the intervals between them can be counted by anyone.
That is the measurement this index reports.
It is a narrower claim than the one we started with, and it has the advantage of being true.
This is the finding the original design would have missed entirely.
CryptoRank's July 2026 research covered 10,110 listings and 4,005 delistings across ten major exchanges, and its methodology section notes that MEXC almost never reports its delistings, so the platform was effectively excluded from the delisting side of that analysis. MEXC does publish delistings.
The monthly ST warning announcements have run continuously through 2026, each carrying a dated table of tagged tokens and their scheduled removal dates, alongside individual notices for specific pairs.
Both statements are accurate, and the gap between them is the point.
Publishing a notice on a web page is not the same as publishing it in a form an independent database can ingest, and a reader relying on third-party trackers rather than the platform's own feed will see an incomplete picture as a result.
The same research reports that OKX delisted no tokens in the first half of 2026, while OKX's own dated notices from January and March 2026 describe removals that ended with withdrawals being closed.
We do not read that as an error on either side.
The likely explanation is that the two are counting different units, with asset-level removals on one side and trading-pair removals on the other.
The practical consequence is the same either way: check the platform's own announcement archive rather than a third-party summary, because the summary may be counting something other than what you are asking about.
Good disclosure is not the same as a good asset.
A platform that tells you precisely what it listed and precisely when it will remove it has still told you nothing about whether the token is worth owning.
The direction of the market also runs against the assumption that listings keep accelerating.
The listing race this index set out to measure is, on the numbers, slowing down.
MEXC operates one of the eight platforms measured here, and this report is published on MEXC Learn.
Three things were done to limit the effect of that.
The six measures were fixed after an earlier timestamp-based design failed, and before the eight-platform data collection began; the failure and the redesign are documented in full above so readers can judge that sequence for themselves.
Every figure comes from a platform's own dated public announcement, which means any reader can check the same pages and reach the same numbers.
Findings that reflect badly on MEXC are reported at the same level of detail as the rest, including the shortest notice period in the index, a withdrawal window at the bottom of the range, and the third-party finding about delisting data.
We publish no overall score, because the useful output here is the underlying dates rather than our opinion about them.
If you hold small positions across several venues and rarely check them, the withdrawal window is the column that matters, and the shortest windows here give you about a month.
If you trade newly listed tokens actively and exit within days, the notice period matters more, because you were never going to be holding at the cut-off anyway.
If you want to know whether a platform's listing pipeline is actually running rather than how large its back catalogue is, the published listing count is the only column that answers that, and one platform fills it.
If you rely on a third-party tracker or an alert channel rather than the exchange's own feed, the verifiability finding above applies directly to you, and the fix is to subscribe to the announcement archive itself.
If you are choosing where to keep a long-term position, none of these columns is the right basis for that decision, and the survival data above matters far more than the paperwork.
Every platform in this index maintains a public announcement archive, and those pages are the only authoritative record of what was listed or removed and when.
Third-party calendars and alert bots are secondary, and as the section above shows, they can be incomplete in ways that are not visible from the summary.
MEXC does not serve US residents, and US traders should use a platform licensed to operate in the United States, such as Coinbase or Kraken.
The disclosure practices described above still apply to whichever venue you use, and the announcement archives referenced here are public regardless of where you live.
EEA readers should factor that into any decision about where to hold assets.
Which exchange lists new coins first?
Decentralised exchanges list first, within minutes, because no review is required.
Among centralised venues, major platforms open within hours of each other for large launches, and no reproducible public measurement ranks them.
Which exchange publishes how many tokens it lists?
MEXC is the only platform in this index publishing a periodic listing count, reporting 135 new tokens in March 2026 at 65% platform-first.
The other seven publish catalogue totals but no periodic figure.
Why is exchange listing speed so hard to measure?
There is no public database of first centralised listing dates that can be queried by quarter, so a neutral sample cannot be built.
Platforms also list large launches simultaneously, which leaves nothing to rank.
Which exchanges publish a dated delisting archive?
All eight platforms in this index do, and each archive is public and readable without an account.
KuCoin additionally maintains a standing table of trading, deposit, and withdrawal closing times for delisted tokens.
How much notice do exchanges give before delisting a token?
The shortest verified interval in this index was 1 day and the longest was 14 days.
Most platforms published their notices six or seven days ahead of the trading halt.
How long do I have to withdraw a delisted token?
Between 28 and 92 days across this index, measured from the trading halt, with Gate setting no cut-off.
Kraken automatically liquidates balances left after its cut-off.
How many new tokens do major exchanges list each quarter?
CryptoRank recorded 351 new listings across ten major exchanges in the second quarter of 2026, the lowest quarterly total since the third quarter of 2023.
Most exchanges do not publish their own listing counts.
Does a fast listing mean the token was reviewed?
No, and speed and review depth generally move in opposite directions.
Around 32% of newly listed tokens across the top 12 exchanges record positive price action immediately after listing, per CoinGecko.
Should I trust a third-party listing tracker?
Use it for discovery and confirm against the exchange's own announcement archive.
Independent databases can miss removals that the platform did publish, as this index documents.
This index measures published disclosure practices, not listing speed, and it does not claim to rank which platform lists tokens earliest.
Notice periods are the shortest interval observed in a sample of dated notices from January to August 2026, which means they describe what happened rather than what is guaranteed.
KuCoin's notice period could not be verified because announcement publication dates were not established from its own dated pages at the time of checking.
Timezone conventions are inconsistent both within and between platforms, with some notices in UTC and others in UTC+8, and all intervals here are calculated in whole days to avoid overstating precision.
Full token removals and trading-pair removals are different kinds of event, and the table notes which is which rather than merging them.
Newly listed crypto assets are among the most volatile instruments in the market, and low circulating supply, concentrated holdings, unlock schedules, and shallow order books can combine to produce losses of most or all of a position within hours.
Nothing here is investment, financial, legal, or tax advice.
Listing, delisting, and withdrawal terms change without notice, so confirm current terms on the relevant platform before acting.