iGaming PSP Comparison 2026: Card vs Crypto-Native Payment Gateways

The iGaming payment stack split into two rails in 2026, and operators picking one are leaving money on the table. Card acquiring still moves the licensed deposit volume that regulators expect, but crypto-native gateways now handle the volumes a card stack would have rejected, repriced punitively, or held in reserve for six months. The question this guide answers is not "Worldpay or Nuvei" or "BitPay or CoinGate". It is "which card acquirer plus which crypto rail matches your volume tier, your license, and your tolerance for held capital".

Eleven providers in our database serve iGaming books at production volume. Five card-side. Six crypto-native. None of them is a "best gateway" in isolation. The sections below shortlist each set, then collapse the choice into a volume-tiered decision rule.

Why operators run two rails in 2026

The April 2026 VAMP threshold drop is the proximate cause. Visa moved the excessive merchant ratio from 2.2 percent to 1.5 percent on April 1 with $8 per disputed transaction and no warning tier, per the Visa fact sheet and Cside's 2026 playbook. iGaming dispute rates run higher than mainstream e-commerce because of buyer's-remorse-after-loss disputes and tighter issuer scrutiny on MCC 7995, so a tighter threshold lands harder on iGaming books than on retail.

Crypto-native gateways do not face the VAMP problem in the same way. The customer's card transaction still flows through Visa or Mastercard, but the merchant of record is the gateway, and the gateway aggregates disputes across its merchant pool. The operator's own MID does not absorb the chargeback burden on volume routed through a crypto-native rail. That architectural difference is why operators are layering crypto-native rails alongside card stacks rather than replacing one with the other.

The second reason is structural cash flow. A card acquirer at $2 million monthly volume holds 5 to 10 percent rolling reserve for six months, which means $600,000 to $1.2 million of working capital sits at the acquirer at any given time. A crypto-native rail settles in USDC or USDT to a wallet, no reserve, no underwriting freeze risk. Operators report the cash-drag math is what moved their boards, not the dispute math.

The third reason is segment fit. Players under 35 in regulated EU markets adopt crypto deposits at materially higher rates than legacy cohorts, per NOWPayments operator data. A card-only stack forces those players through worse approval rates, while a crypto rail picks them up at instant settlement and ~1 percent fees.

Five card acquirers worth a shortlist

These five all underwrite iGaming MIDs at production volume in 2026 and have actual operator references in regulated markets. The first comparison sets the dimensions; the paragraphs below it explain the fit.

Worldpay runs prebuilt integrations with SoftSwiss, EveryMatrix, and OpenBet, and its underwriting handles UKGC- and MGA-licensed books at production volume. Settlement is T+7 or longer, deposit fees blend 1.5 to 3.5 percent. Reserves at 5 to 10 percent are the negotiable starting point, not the floor. The trade-off operators report is slower settlement and a six-month migration cost to switch out once embedded.

Where Nuvei wins is iGaming-specific tooling. Network tokenization, 3DS exemption flows, and a published VAMP playbook addressing the April 2026 threshold change. Pricing is custom in the same 1.5 to 3.5 percent band as Worldpay; settlement is T+7 or longer. The differentiator operators reference is approval-rate uplift on MCC 7995 versus generic acquirers, sourced from Nuvei's own benchmarks rather than independent measurement.

Paysafe operates 30+ local payment methods alongside card acquiring, plus the digital-wallet brands Skrill, Neteller, and paysafecard under one roof. That matters for operators whose player base in regulated EU markets already uses those wallets. Settlement is T+2 to T+3, faster than Worldpay or Nuvei direct. Fees blend 1 to 2.9 percent on custom contracts.

IXOPAY's value is portfolio diversification. The product sits in front of multiple acquirers and routes per transaction by issuer BIN, decline patterns, or fee. The orchestration fee is 0.1 to 0.5 percent on top of the underlying acquirer cost. The reason to pick IXOPAY is failover: if one acquirer rate-jacks or pulls out of a vertical, the orchestration layer routes around it without a six-month migration project.

Solidgate settles at T+1 on certain rails, faster than any other card-side option here, and charges 0.3 to 0.8 percent orchestration fees on top of acquiring. The acquirer roster is smaller than IXOPAY's, so the redundancy story is weaker. Operators who pick Solidgate prioritize cash velocity over the broader acquirer pool.

Six crypto-native rails worth a shortlist

These six all settle Visa and Mastercard deposits to a stablecoin wallet, eliminate the rolling reserve mechanic, and target the operators that either cannot get card-only approval or want a parallel rail to reduce acquirer dependency.

BitPay opened in 2011 as the first crypto payment processor of any scale and remains the regulated US-friendly option in 2026. U.S.-licensed money transmitter coverage across 40+ states, T+1 fiat conversion when the merchant wants off-ramp, around 1 percent fees. Chain support is conservative (Bitcoin and Ethereum first, others added selectively), which fits operators that want a regulated counterparty more than maximum coin breadth.

CoinGate holds a Lithuanian VASP registration under the EU AMLD framework and runs Chainalysis-grade KYT on inbound wallets. Instant settlement, around 1 percent fees, broader chain support than BitPay. A reasonable default for European-licensed operators who want a crypto rail without taking on offshore counterparty risk.

Non-custodial fiat settlement is what differentiates CoinsPaid. The merchant receives funds without CoinsPaid ever holding the fiat, which insulates the operator from the freeze-risk custodial gateways carry. Around 0.8 percent fees, instant settlement. CoinsPaid's customer roster includes 1xBet, 1xBit, and other major EU-facing casino brands.

Where NOWPayments competes is breadth. 350+ cryptocurrencies supported, 0.5 to 1 percent fees, instant settlement. Custodial-by-default with a non-custodial option, which is the structural trade-off relative to CoinsPaid. The fit is operators serving global player bases who need long-tail altcoin support alongside USDT and USDC.

Triple-A's Singapore MAS license is what banking partners care about. A Major Payment Institution license from the Monetary Authority of Singapore is one of the stricter crypto-payment regimes globally, per Triple-A's own licensing documentation. Supported coins are more conservative than NOWPayments or CoinPayments, but the Singapore-bank fiat off-ramp is the cleanest in the segment for APAC operators or anyone wanting a tier-1 regulator behind the gateway. 0.5 to 1 percent fees, instant settlement.

CoinPayments handles the long tail. 350+ coins, 0.5 to 1 percent fees, instant settlement, lighter compliance documentation than the four above. Production fit for operators in gray markets or those running primarily on USDT volume. Less appropriate as the only rail for a regulated EU operator.

The decision rule per volume tier

The "best" combination is the one whose constraints match yours. Run the rule against your monthly card volume and your license stack; the answer comes out in one row.

1. Under $500k monthly card volume, regardless of license: orchestrator (IXOPAY or Solidgate) plus a crypto-native fallback (CoinGate or NOWPayments). Direct acquirer relationships are overkill at this scale, and the orchestration fee is cheaper than maintaining two acquirer accounts.
2. $500k to $2 million monthly with a UKGC, MGA, or DGOJ license: Worldpay or Paysafe direct as primary, with IXOPAY layered for failover, plus CoinGate or CoinsPaid as a crypto rail. The card stack carries the regulator-facing volume; the crypto rail picks up segments where card approval is poor.
3. $2 million to $10 million monthly with a tier-1 license: Nuvei or Worldpay direct, IXOPAY for redundancy, plus Triple-A or CoinsPaid for non-custodial crypto settlement. At this volume the rolling-reserve drag is a board-level number, so non-custodial crypto matters more than coin breadth.
4. Above $10 million monthly: dual card acquirers (Nuvei and Worldpay) on IXOPAY orchestration, plus two crypto rails for redundancy (CoinsPaid for non-custodial volume, NOWPayments for breadth on retention segments). Concentration risk is the constraint, not cost.
5. Offshore-licensed operators (Curacao GCB, Anjouan, Costa Rica): card acquiring is mostly closed. Run a crypto-first stack: CoinsPaid or Triple-A for non-custodial, NOWPayments or CoinPayments for breadth. Add Paysafe selectively if you can clear their post-2024 Curacao reform underwriting.

What to negotiate before signing

The rate sheet is not the contract. The clauses below decide whether month-nine economics match the day-one quote.

Card-side levers

• Push on the rolling reserve percentage and hold period. Ten percent for six months is the negotiable starting point, not the floor. Get the release schedule in writing on a fixed monthly cadence, not "at acquirer discretion".
• Cap chargeback fees at $25 to $35 per dispute. Some acquirers quote $40 plus, and the math compounds quickly on iGaming dispute volumes.
• Clarify who carries MID concentration risk if the acquirer pulls out of iGaming mid-contract. The answer decides whether you need a second acquirer in parallel or whether the orchestrator covers the gap.
• Lock in the network-token rollout timeline. Approval-rate uplift from tokens runs in the 4 to 6 percent band on MCC 7995 per Visa's PERC issuer-side data, but only after the rollout actually ships. Vague timelines are a tell.

Crypto-side levers

• Decide on custody upfront. Custodial gateways carry the same freeze-risk shape as card acquirers; non-custodial gateways do not. Ask what specifically triggers a freeze in the custodial model. Few providers will commit the trigger conditions to writing.
• Confirm fiat off-ramp restrictions per geography before signing. The wallet-side coin support is rarely the constraint; the fiat-out side is.
• Confirm KYT coverage on inbound wallets, especially for operators running USDT volume given the FinCEN GENIUS Act rule changes from April 2026.

The two-rail pattern is not a hedge. It is the structural answer to a 2026 stack where card acquirers handle the volume their underwriting wants and crypto rails handle the rest. Pick one card acquirer or orchestrator from the shortlist above. Pick one crypto-native gateway whose compliance posture matches your jurisdiction. Run both in production for ninety days, measure approval rates, settlement velocity, and held-capital cost. The combination your numbers identify is the right one for your book.