Normal view

There are new articles available, click to refresh the page.
Today — 14 September 2026Main stream

Crypto On Ramp Fees: Why More Volume Does Not Fix Your Margins

By: Fystack
14 September 2026 at 06:57

In the deposit address model most ramps run, each new address takes two on-chain transactions before it returns anything. Here is what a ramp transaction costs end to end, and which part an operator can still change.

Crypto On Ramp Fees: Why More Volume Does Not Fix Your Margins

TL;DR

In the deposit address model most ramps run, each new address takes two on-chain transactions before it returns anything. Here is what a ramp transaction costs end to end, and which part an operator can still change.

Add a thousand customers and you add a thousand deposit addresses. In the standard model, each one takes two transactions before a cent of it reaches your hot wallet.

An engineer at a crypto exchange added his up. The answer came to roughly 270 million gas a month, at 50 deposits a day.

  • 50 deposits/day times 2 transactions/deposit = 100 transactions/day
  • 100 transactions/day times 30 days/month = 3,000 transactions/month
  • 3,000 transactions/month times ~90,000 gas/transaction ~270M gas/month

Why Deposit Wallets Cost Two Transactions Each

A fresh deposit address arrives empty. It holds the customer’s USDT but none of the native token needed to move it, so the platform funds the address first and sweeps it second. Both transactions are paid by the platform, on every deposit.

This is why volume does not rescue the number. Every extra customer brings an extra funding transaction with them, so the cost climbs at the same rate as the customer count while revenue per customer stays flat. That is one line on a longer bill.

Why Banking Costs Cannot Be Negotiated

Frax CEO Sam Kazemian quoted his JPMorgan banker word for word in 2024: “we have to close anyone’s account that we know their primary source of income/wealth is crypto. This is directly from the top from Jamie.”

n December 2025 JPMorgan froze accounts belonging to Blindpay and Kontigo over disputed transactions.

Licensing has not changed the pattern. 338 firms now hold full MiCA authorisation across 30 EEA markets, up from around 60 at the start of the year, and banks still decline them.

MiCA carries no provision requiring a bank to serve a licensed CASP, while the European Banking Authority continues to treat the sector as enhanced due diligence whatever an individual firm has cleared. Around 86 percent of European crypto companies report repeated account closures.

An operator carrying a 10 percent reserve has no lever to pull on any of this, which sends them back to the last row of the table.

How TRON Prices New Deposit Addresses

TRON holds more than 51 percent of all USDT in circulation, so a ramp settling in USDT is almost certainly running there.

TRON charges by the address. One published comparison put a transfer to a brand new address at 27.25 TRX, against 13.4 TRX to one that had been received before, which puts every customer’s first transfer on the expensive path.

Justin Sun’s Proposal #104 post confirms the energy unit price was halved in August 2025, the largest cut in TRON’s history. The same post commits super representatives to quarterly fee reviews, so an operator’s per transaction cost gets decided at a governance vote they do not attend.

TL;DR

In the deposit address model most ramps run, each new address takes two on-chain transactions before it returns anything. Here is what a ramp transaction costs end to end, and which part an operator can still change.

Add a thousand customers and you add a thousand deposit addresses. In the standard model, each one takes two transactions before a cent of it reaches your hot wallet.

An engineer at a crypto exchange added his up. The answer came to roughly 270 million gas a month, at 50 deposits a day.

  • 50 deposits/day times 2 transactions/deposit = 100 transactions/day
  • 100 transactions/day times 30 days/month = 3,000 transactions/month
  • 3,000 transactions/month times ~90,000 gas/transaction ~270M gas/month

Why Deposit Wallets Cost Two Transactions Each

A fresh deposit address arrives empty. It holds the customer’s USDT but none of the native token needed to move it, so the platform funds the address first and sweeps it second. Both transactions are paid by the platform, on every deposit.

This is why volume does not rescue the number. Every extra customer brings an extra funding transaction with them, so the cost climbs at the same rate as the customer count while revenue per customer stays flat. That is one line on a longer bill.

Why Banking Costs Cannot Be Negotiated

Frax CEO Sam Kazemian quoted his JPMorgan banker word for word in 2024: “we have to close anyone’s account that we know their primary source of income/wealth is crypto. This is directly from the top from Jamie.”

In December 2025 JPMorgan froze accounts belonging to Blindpay and Kontigo over disputed transactions.

Licensing has not changed the pattern. 338 firms now hold full MiCA authorisation across 30 EEA markets, up from around 60 at the start of the year, and banks still decline them.

MiCA carries no provision requiring a bank to serve a licensed CASP, while the European Banking Authority continues to treat the sector as enhanced due diligence whatever an individual firm has cleared. Around 86 percent of European crypto companies report repeated account closures.

An operator carrying a 10 percent reserve has no lever to pull on any of this, which sends them back to the last row of the table.

How TRON Prices New Deposit Addresses

TRON holds more than 51 percent of all USDT in circulation, so a ramp settling in USDT is almost certainly running there.

TRON charges by the address. One published comparison put a transfer to a brand new address at 27.25 TRX, against 13.4 TRX to one that had been received before, which puts every customer’s first transfer on the expensive path.

Justin Sun’s Proposal #104 post confirms the energy unit price was halved in August 2025, the largest cut in TRON’s history. The same post commits super representatives to quarterly fee reviews, so an operator’s per transaction cost gets decided at a governance vote they do not attend.

Three Ways Ramps Reduce Sweeping Costs

Deterministic addresses with proxy contracts

The engineer from the opening rebuilt his sweeping on CREATE2 with ERC-1167 minimal proxies and, in that one implementation, cut gas consumption by 84 percent

Threshold sweeping

Small balances stay put until the amount clears the cost of moving it.

Fystack runs this as sweep tasks set on a USD value, with one task attached across many deposit wallets so a threshold change lands everywhere at once, and a gas tank keeping those wallets funded for the transfer.

// Create wallet with new sweep task
await sdk.createWallet({
name: 'User Payment Wallet',
walletType: WalletType.Standard,
sweepTaskParams: {
minTriggerValueUsd: '100.00',
destinationWalletId: '123e4567-e89b-12d3-a456-426614174001',
destinationType: 'internal_wallet'
}
})

Energy acquired instead of burned

Staking TRX or renting from an energy pool both avoid the burn. Guarda published a comparison of 11 USDT transfers costing $76.79 burned against $17.96 rented.

A TRON DAO ambassador puts the unit economics near 5.5 TRX per 100,000 energy, enough for two USDT transfers, with the deposit returned when the rental ends.

Rental turns the per transfer cost into a working capital decision:

A TRON withdrawal in Fystack with energy renting active: 2.97 TRX rented against 24.03 TRX burned, roughly $6.96 saved on one transfer.

Where to Start Cutting Sweeping Costs

Sweeping is the one cost line still open to a decision, and the spread between a naive build and a considered one runs wide enough to matter at ramp margins.

If deposit wallets are multiplying faster than revenue and nobody owns the sweeping bill, that number is worth pulling before the next volume step.

Fystack can help there: stablecoin custody and payment infra, self hosted, with threshold sweeping and TRON energy renting built in. The signing core, mpcium, is open source.


Crypto On Ramp Fees: Why More Volume Does Not Fix Your Margins was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.

What Crypto Cards Reveal About the Future of Payments

By: Fystack
14 September 2026 at 06:56

Crypto card monthly volume grew 106% annually from early 2023, driven almost entirely by programs that run on Visa. Mastercard is running two separate strategies at once: consumer card agents and a B2B settlement infrastructure acquisition. The card works as a bridge for existing merchants. For new ones with no entity and no processing history, protocols like x402 are the only option that works today

TL;DR

Crypto card monthly volume grew 106% annually from early 2023, driven almost entirely by programs that run on Visa. Mastercard is running two separate strategies at once: consumer card agents and a B2B settlement infrastructure acquisition. The card works as a bridge for existing merchants. For new ones with no entity and no processing history, protocols like x402 are the only option that works today

What Crypto Cards Reveal About the Future of Payments

Introduction

Artemis Research reported in January 2026 that total crypto card monthly volume grew from $100 million in early 2023 to over $1.5 billion by late 2025, a 106% compound annual growth rate. Annualized, the market reached $18 billion — close to the $19 billion in peer-to-peer stablecoin transfers over the same period.

By July 2026, Paymentscan tracked $691.7 million in a single month from on-chain settled card programs alone, with $8.094 billion in cumulative volume across 128.7 million transactions and 467,000 wallet addresses.

How Much People Are Spending on Crypto Cards

The biggest program by monthly volume is RedotPay at roughly $380 million per month. KAST ranks third. Both run on Visa. Paymentscan’s July 2026 data shows Visa at 98.5% of on-chain crypto card volume. Mastercard at 1.5%.

Visa built its position by partnering early with infrastructure providers like Rain and Reap, which issue cards for many downstream programs at once. One Visa infrastructure deal covers many products. Mastercard built its crypto card presence through direct partnerships with centralized exchanges — Bybit, Revolut, Gemini — which ties its volume more closely to exchange trading cycles.

How Visa and Mastercard Compete on Crypto

Both networks are extending card infrastructure to AI agents. Noah Levine at a16z noted in March 2026 that Mastercard Agent Pay is live for US cardholders and Visa’s Intelligent Commerce framework is in pilot. The Agentic Commerce Protocol, built by Stripe and OpenAI, has Etsy live with over one million Shopify merchants set to follow.

At the consumer layer, both networks are converging on the same approach. At the settlement layer, they are not. Mastercard announced in March 2026 that it would acquire BVNK for up to $1.8 billion, including $300 million in contingent payments. BVNK processes $30 billion annually for clients including Worldpay, Deel, and Rapyd. The stated goal: 24/7 stablecoin settlement for processors and acquirers, and stablecoin checkout within Mastercard’s payment gateway.

How Crypto Cards Work as a Bridge

A crypto card does one job: it converts a wallet balance into what a merchant already accepts. The cardholder pays from USDC. The merchant receives a standard card transaction. Artemis confirmed that the vast majority of crypto card transactions today settle in fiat at transaction time, so the merchant sees only local currency regardless of which stablecoin the user holds.

That job has real reach. RedotPay users in Southeast Asia, LATAM, and Africa can pay at any Visa-accepting terminal without the merchant needing to know anything about crypto.

But the card cannot serve every merchant. Levine at a16z put it directly:

“Cards serve every merchant a processor can underwrite. Stablecoins serve every merchant a processor cannot.”

A developer who builds an API tool in a weekend has no legal entity, no website, and no chargeback history — a profile that is very difficult to underwrite. Levine’s conclusion: “These merchants will not be choosing stablecoins over cards. They will be choosing stablecoins over nothing.”

What AI Agents Need That Cards Cannot Provide

In March 2026, a marketplace at mpp.dev opened with over 60 services built for AI agents: SEC filing search charges per query, image generation at fractions of a cent, physical letters printed and mailed from a document and address. The protocol powering it is MPP (Machine Payments Protocol), built by Stripe and Tempo, which lets agents pay using cards, stablecoins, or Lightning in a single HTTP request.

In its first week, 894 agents made 31,000 transactions at prices from $0.003 to $35 per request, according to a16z. No checkout page. Pricing embedded in the HTTP response. The agent reads the schema, pays, receives output.

Two protocols now make this model possible at the infrastructure level. x402 embeds stablecoin payments directly in HTTP requests with no merchant account or processor onboarding. MPP supports cards, stablecoins, and Lightning in the same request, with the mpp.dev marketplace as its live implementation. As a16z noted, “x402 and MPP each take a different approach, but both embed payments directly into HTTP requests.”

MoonPay’s PayBox handles the full stack for agents — MPC key sharding in a trusted execution environment, policy limits before any transaction, virtual cards scoped per merchant for fiat rails when needed.

Conclusion

The card worked because neither side had to change anything. As Levine, a16z, wrote: “The next generation of merchants won’t have storefronts. They’ll have endpoints.” The card serves that well for the merchants that already exist. The infrastructure being built now is for the ones that do not.

About Fystack

Fystack is an enterprise-grade, self-hosted MPC custody platform for fintech teams and crypto businesses. The core signing infrastructure, mpcium, is open-source. Fystack supports multi-chain wallet operations across TRON, ETH, BNB, Solana, Polygon, and more, with a policy engine that enforces spend rules before any signing happens.

If you are building payment infrastructure that involves automated signing, wallet custody, or agent payment flows, Fystack has the full product overview. The policy engine source is on GitHub.


What Crypto Cards Reveal About the Future of Payments was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.

❌
❌