Crypto Payment Gateways for Hard-to-Place Merchants

Merchants in difficult categories usually solve acceptance twice. The first solution is a specialist card processor with higher rates, a rolling reserve and deeper underwriting. The second is a crypto rail, which removes the reserve and the chargeback entirely and replaces them with a different set of problems. This piece is about the second solution and where it does and does not fit.

What the crypto rail actually removes

Three things disappear from the merchant's economics.

The rolling reserve goes first. Specialist card acquirers hold a percentage of settled volume against future disputes, typically for months. A confirmed on-chain payment cannot be reversed by the payer, so there is nothing to reserve against. For a merchant on a double-digit reserve percentage, that working capital release is usually a larger number than the entire annual processing fee.

The chargeback goes with it. There is no scheme mechanism for the customer to reverse a settled transaction. Refunds still exist — they are simply new payments the merchant chooses to send — but they become a policy decision rather than an obligation imposed by a card network and adjudicated by someone else.

The category-based surcharge goes third. Card pricing for difficult verticals reflects dispute exposure that crypto rails do not carry, which is why crypto processing rates in this space cluster well below specialist card rates rather than above them.

What arrives in exchange: confirmation latency instead of instant authorisation, price movement between quote and settlement unless the merchant converts, and on-chain screening that can reject an incoming payment before it ever lands. Those three are the real subject matter, and most published comparisons skip all of them.

What it does not remove

Licensing, first. Crypto acceptance does not substitute for the authorisations a business needs to operate. Regulated verticals are still regulated, and a serious provider will decline a merchant that cannot evidence its own permissions — Speend, for instance, requires an active gaming licence from an accepted jurisdiction before onboarding an iGaming operator, and turns down operators without one.

Payouts, second. Deposits and withdrawals are different problems with different controls, and a provider that handles one well may handle the other poorly. Ask about both.

Banking, third. Converting crypto revenue into fiat still touches the banking system, and that is where hard-to-place merchants meet the same friction they were trying to route around. Merchants that settle and hold in stablecoins avoid the question; merchants that need fiat in an account do not.

Speend

Speend runs merchant-side verification: the business passes KYB, and the payer pays from any wallet without opening an account. Pricing starts at 0.5% with no setup charge, no monthly fee, no minimum turnover, and network costs passed through at blockchain cost rather than marked up. Volume tiers move the rate down, and the tier for licensed iGaming operators starts at 0.2%.

Four merchant categories have dedicated coverage rather than a generic high-risk bucket: iGaming, ecommerce, SaaS and subscription businesses, and adult content platforms. For adult specifically there is no high-risk surcharge and no chargeback exposure by construction, which is the entire reason the category is difficult in card acquiring. For a business in that vertical, an adult payment gateway built on crypto rails prices the same as one built for mainstream retail.

Onboarding is 1–3 business days for KYB in the general case and 24–48 hours for licensed iGaming operators, with technical integration quoted at 24 hours and support answering in 10–15 minutes including weekends. Asset coverage runs to 300-plus coins across 18 networks, with USDT open on five rails — Tron, Ethereum, BNB Smart Chain, Polygon and Solana — and Bitcoin available on mainnet, Lightning and SegWit.

The controls are the part that matters for a difficult category, because a provider that cannot evidence them will eventually create a banking problem for its merchants. Speend runs MPC wallets with no single point of key assembly, role-based access with a per-operation audit log, separation of hot and cold wallets with the treasury held cold, on-chain screening on every incoming transaction with flagged funds rejected automatically, 24/7 monitoring with automatic alerts, a sandbox that mirrors production one-to-one, and a 99.95% availability SLA on geographically distributed infrastructure. Settlement can hold in crypto, auto-convert into a stablecoin at the moment of receipt, or withdraw to a self-custodial wallet, an exchange or a supported banking channel. For merchants who need crypto payment processing that does not price their category as a penalty, that combination is the case.

B2BINPAY

Built for exactly this end of the market: forex and CFD brokers, exchanges, licensed gaming operators and financial institutions. Volume-banded pricing that reaches the low end of the category at scale, wallet-as-a-service, and white-label orchestration for businesses that want their own brand on the checkout. There is an onboarding charge, and the platform does not cover US merchants.

NOWPayments

Non-custodial, which removes the question of whether the provider is sitting on your revenue. Fast self-serve onboarding, 350-plus assets, low base rate with verification-linked discounts. Fiat settlement relies on third-party integrations, and the company holds no MiCA, PI or EMI authorisation at the time of writing — which is a live consideration for merchants whose own banking partners ask about counterparty licensing.

BTCPay Server

Self-hosted, open-source, no company in the middle and no acceptable-use policy to be declined under, because there is no one to decline you. The merchant runs the node and holds every key; the cost moves from a percentage to engineering time. Realistic only for businesses with someone to own the infrastructure.

Providers to check carefully, whatever the marketing says

This is the section that separates a usable shortlist from a liability, and it is not optional in difficult categories, because a merchant's provider becomes part of the merchant's own risk profile the moment a banking partner starts asking questions.

Regulatory action in this sector is documented and public. Canada's financial intelligence unit issued an administrative penalty of C$176,960,190 against a crypto payment operator on 16 October 2025 over reporting failures, the largest penalty that agency has published. Separately, blockchain analytics firms track continuity between sanctioned platforms and their successor services, including cases where a payment processor's parallel brand received its opening liquidity directly from a platform that had been the subject of an international law enforcement operation — TRM Labs documented that pattern across several 2025 rebrands.

Screening exposure is measurable too. Chainalysis recorded more than $820 million in on-chain ransomware payments during 2025, an 8% decline year on year, and enforcement over that period increasingly targeted infrastructure and payment layers rather than only the groups themselves. A merchant whose provider does not screen incoming transactions inherits that flow.

Practical checks before shortlisting: confirm the operating entity and where it is authorised, look for published enforcement actions against that entity or its parent, ask which analytics vendor performs the screening, ask what happens to a flagged payment operationally, and check whether the brand is a recent relaunch of something else. Any provider that will not answer the first three in writing has answered them.

What to negotiate

Ask for the KYB timeline and the integration timeline as separate numbers. Ask whether network fees are billed at cost or with a spread. Ask what the volume tiers are and whether they carry a committed minimum, because a tier that requires a commitment is a different product from one that does not. Ask how payouts are handled, including limits and manual review triggers, since that is where withdrawal-heavy businesses discover the constraints. And ask for a sandbox that behaves like production, then test underpayment, overpayment and a wrong-network send before you go live rather than after.

FAQ

Why do card acquirers decline these merchants?

Dispute exposure. Categories with elevated chargeback rates cost the acquirer money after the sale, so pricing carries a surcharge and settlement carries a reserve. Some acquirers decline the category outright rather than price it.

Does accepting crypto remove the need for licensing?

No. Payment rails and operating permissions are separate questions. Serious providers verify that a merchant holds whatever authorisations its business requires, and decline merchants that cannot evidence them.

What is a rolling reserve and does crypto have one?

A percentage of settled volume held by the acquirer against future disputes, usually released after a set period. Crypto rails have no equivalent, because settled on-chain payments cannot be reversed by the payer.

Can a merchant still issue refunds?

Yes — as new outbound transactions. The difference is that the merchant decides, rather than a scheme rule deciding for them. Refund policy should be written down before launch, because customers will ask.

What happens if an incoming payment is flagged by screening?

It depends on the provider. Some reject before crediting the balance, some freeze pending review. The operational question — whether the order stays open and who tells the customer — is worth answering during integration rather than during the first incident.

Is stablecoin settlement safer than holding volatile assets?

For price exposure, yes. It substitutes market risk for issuer and regulatory risk, which is why merchants with mixed international customer bases increasingly accept more than one stablecoin rather than standardising on one.