Top Crypto Market Makers for Token Launches

Choosing a crypto market maker is not a vendor selection. It is a decision about which set of incentives sits on the other side of your order book for the next twelve months, and the contract structure decides that before any firm's reputation does. Two projects can hire the same desk, on the same exchanges, in the same month, and get opposite outcomes. One signed a retainer. The other signed a token loan with a call option attached.

The compensation model is the product

Three structures dominate term sheets, and hybrids of them are increasingly common. Under a retainer, you pay a monthly fee in cash and supply your own inventory; the desk is contracted to hold a spread, a depth and an uptime, and its revenue does not move with your token price. Under a token loan with a call option, you lend tokens as trading inventory and pay no cash, and the desk is compensated by an option to buy some of those tokens later at a preset strike. Under profit-share, the fee is replaced by a split of realised trading P&L. Flowdesk publishes a clear description of the first two, which is more than most firms in this market do.

The loan-and-call structure is not automatically predatory. For a large project with deep liquidity it is usually a reasonable way to buy market making without spending cash. It does contain two incentives that are easy to run together, and they are not the same thing.

The first is the option. It pays out above the strike, so that part of the desk's compensation is tied to your price arriving at a number rather than to your book being orderly. The second is a trading position: the desk holds borrowed inventory it must eventually return, so selling into demand and repurchasing lower is profitable on its own terms, whatever the option does. Which path is attractive depends on the strike level, the size of the option, the loan-return conditions, any trading restrictions, and the termination provisions. A loan-and-call deal is not one incentive. It is a set of them, and the terms decide which dominates. Cointelegraph has documented launches where the selling path won.

Which brings the practical difficulty. The terms are almost never disclosed: loan size, strike and repayment conditions stay between the two parties (either because the desk prefers it that way, or because nobody on the project side thought to ask). So your holders cannot see the incentive acting on your price, your own community cannot, and in most cases neither can an article like this one. That is a hard limit on what any comparison can tell you, and it is why the table further down ranks on disclosure rather than on size.

There is one place the terms do surface. Where a DAO hires a desk, the treasury has to vote, so the deal gets published. GSR's proposal to Stake DAO set out a $100,000 setup fee, a $20,000 monthly retainer and a $1 million BTC and ETH loan. If you want to see what real terms look like before negotiating your own, the governance forums of DAOs that have hired a desk are the most reliable public source available.

We have written about this structure before, with a worked example, in our piece on DEX, CEX and the role of market makers. The position there has not changed.

What the enforcement record actually shows

A badly structured loan-and-call deal misaligns a legitimate service. What follows is something else: outright fraud sold as market making. The two are related in that both turn on a project buying something other than liquidity, but the second does not follow from the first, and nothing in the public record shows that desks using loan-and-call are more likely to commit it.

On 9 October 2024 the SEC charged three firms describing themselves as market makers, Gotbit, ZM Quant and CLS Global, along with nine individuals. The DOJ brought parallel criminal charges and added a fourth firm, MyTrade. The allegation was wash trading: self-dealing that produces volume and no price discovery, run by bots that at times generated billions of dollars of artificial volume a day.

The investigation is worth understanding for how it was run. The FBI created its own token, NexFundAI, and offered it to the firms. The resulting case, Operation Token Mirrors, seized more than $25 million. Nobody involved was confused about what a market maker does. Volume was the deliverable, and the projects buying it knew that.

Liquidity and volume are different products. Volume makes an exchange listing tier look achievable and a chart look alive; liquidity lets a real buyer fill a real order without moving the price. A desk that treats them as the same thing has told you which one it sells.

The firms worth a conversation

Two filters produce the five below. The first is launch relevance: a desk that quotes a listing-stage token with $200,000 of daily volume, not one whose business is institutional flow. That rules out Cumberland, B2C2 and Flow Traders, all of them serious firms and none of them built for the job on this page. The second is disclosure: something material about how the firm operates has to be publicly checkable, whether that is a compensation model, a regulatory registration, a reporting standard or a published term sheet. The second filter is deliberately unkind to the firms that say least, and it is the only one this article can honestly apply, because the variable that matters most is contractual and mostly private.

It is not a ranking, and every figure below is the firm's own published claim.

GSR is the only firm here whose actual terms you can read before you call them. Its Stake DAO proposal is public, it holds a FINRA broker-dealer registration, and it carries derivatives and advisory capability most desks do not. That combination usually prices accordingly. If you are choosing partly on how much you can verify in advance, it is the shortest path to an informed conversation.

Wintermute runs the largest algorithmic book in the sector and the deepest DeFi coverage, and both are genuinely hard to replicate. The question for a launch is whether your token is interesting to it. A desk built for $15 billion of daily flow assigns its attention accordingly, and the mismatch typically shows up as a junior trader and a book that is technically compliant and practically empty.

Keyrock publishes what it reports to clients, which is rarer than it sounds. KPI dashboards and defined reporting turn a retainer into something you can dispute, and a contract you can dispute is the only kind worth signing. If your main worry is that you will not be able to tell whether the desk is doing its job, start here.

Flowdesk publishes its compensation models, including the loan-and-call mechanics that most firms would rather discuss privately. That is a small thing and I read it as a large signal: a desk willing to explain how it gets paid before you ask is a desk with less to manage in the explaining. Retainer-led, multi-venue, and straightforward about which structure it is proposing.

EchoTrade works across more than 90 centralized and decentralized exchanges, with traders holding client order books around the clock against the depth, spread and uptime standards exchanges monitor after listing. It also covers exchange listing support, compliance review and treasury building, and is an official Liquidity Partner of MEXC. It works on a monthly retainer only: no token loan, no call option, no profit share. By the argument this article opens with, that is the cleanest of the three structures, and it removes the question you would otherwise have to ask first.

DWF Labs is active in launch-stage deals and is deliberately not on the list. In 2024 the Wall Street Journal reported that Binance's internal market-surveillance team had investigated the firm after complaints from competing market makers, concluding it had run extensive wash trading during 2023: more than $300 million of it, across Yield Guild Games and at least six other tokens. The firm operates at real scale and some projects will still choose it. Saying that out loud is more useful than a quiet omission, and if you do put it on your own shortlist, that report is the first thing to raise and get answered.

None of these five is the right answer on its own. Where your liquidity should sit in the first place, across centralized venues, decentralized ones, or a market maker, is a prior question, and answering it narrows the shortlist faster than any comparison table.

Unlock schedules are where the two disciplines meet

Every vesting cliff a token design contains is a date on which new supply arrives in a market that somebody has to absorb. Your emission schedule and your market making agreement are two halves of one problem, and they are almost always negotiated by different people at different times without either seeing the other.

The screening arithmetic is simple and rarely done. Take the tokens unlocking in a given month, divide by the trading days in it, and compare that to your daily traded volume. Where a cliff releases supply worth several days of volume at once, the book is being asked to absorb more than the market normally trades, and a spread will be expensive to hold or will not hold at all.

Treat that as a screen and not a verdict, because both sides of it are approximations. Tokens unlocked are not tokens sold; recipients vest into holding about as often as into selling. Reported volume is not absorptive capacity, since some of it is wash-inflated and the rest is not all executable at a price you would accept. What actually decides the outcome is expected net sell flow against genuine executable depth at tolerable slippage. So the ratio flags which dates deserve a proper model. It is not the model. The responses are all upstream anyway: flatten the cliff into a linear release, move the date, or size the position with the desk in advance so the inventory exists before it is needed.

Ask a prospective market maker about your unlock schedule on the first call, because the answer is diagnostic. A desk that asks for the schedule before quoting is pricing the actual risk. A desk that quotes on exchange count and monthly fee alone has not looked at what it is being asked to absorb, and will discover it at the cliff.

How to evaluate a market maker

Five questions, in order of how much they tell you.

• Which compensation model, and why this one for us? A firm that offers only loan-and-call, or cannot explain why it suits your situation, has told you what it optimises for.
• If there is an option, what is the strike and who set it? A strike defined as a premium over the first few days of trading rewards a low opening price. Worth knowing before you sign rather than after.
• What are the contractual KPIs and how are they reported? Spread, depth at defined price bands, uptime, measured and sent to you on a schedule. If those three terms are not yet second nature, what each one actually measures is worth ten minutes before the first call. Without numbers there is no performance to dispute.
• What happens at unlock dates? See above. The quality of this answer separates firms quickly.
• What are the exit terms? Notice period, how loaned inventory is returned, and what happens to the option if you terminate early. This clause is where a bad relationship becomes an expensive one.

What to put in the contract

Four provisions do most of the protective work, and none of them is unusual enough to be a difficult ask.

Put the KPIs in the agreement rather than the pitch deck, with defined measurement: spread at a stated size, depth within stated bands, uptime as a percentage over a stated window. Require reporting you can audit, meaning the desk's own trade data on a fixed cadence rather than a monthly summary. Where an option exists, cap the inventory it can be exercised against and disclose the strike to whoever governs the treasury. And set a notice period short enough to be a real threat: a twelve-month lock with no performance-based exit removes the only leverage you have once the book is not being held.

I would not expect any of this to make a market maker align with your project automatically. It makes the misalignment visible and priced, which is the most contract design can do. The rest is choosing a desk whose compensation does not depend on your token behaving in a particular way, then checking monthly that the book looks like the agreement said it would. Sign it in the right order, too: liquidity is arranged before listings, not after them.

If you are designing the emission schedule these agreements have to absorb, that is the work we do: tokenomics consulting built on modelled and stress-tested designs rather than templates.

Our case studies and the wider portfolio are public, including the token economies behind several of the launches these structures were negotiated around.