The crypto industry has spent a decade promising to make digital assets spendable. The reality has been clunky: sell on an exchange, withdraw to a bank, then spend. That friction is now collapsing. In the last 90 days, a new generation of self-custody crypto cards has emerged that lets users spen...
"The goal is for the line between onchain and offchain to fade away entirely." — Gal Eldar, MetaMask Product Lead
The crypto industry has spent a decade promising to make digital assets spendable. The reality has been clunky: sell on an exchange, withdraw to a bank, then spend. That friction is now collapsing. In the last 90 days, a new generation of self-custody crypto cards has emerged that lets users spend directly from their own wallets — no exchange, no pre-loading, no custodial handoff — at 150 million merchants worldwide.
The MetaMask Card, backed by Mastercard and launched across 49 U.S. states on February 26, is the highest-profile entry. But it is not alone. Gnosis Pay runs on Visa from a Safe smart-contract wallet. Phantom's CASH card on Solana is in early U.S. access. And the infrastructure beneath them — Stripe's Bridge platform, M0's issuance protocol, Monavate's card rails — has quietly matured into a full-stack stablecoin-to-fiat pipeline. Crypto card spending has hit an $18 billion annualized run rate, up from effectively zero three years ago. The self-custody segment is the fastest-growing slice.
This is not another crypto payments hype cycle. The difference this time is structural: regulated bank issuers (Cross River Bank, FDIC-insured), institutional card networks (Mastercard, Visa), and real stablecoin volume (Bridge's throughput quadrupled in 2025) underpin these products. For the first time, the plumbing works.
Crypto card spending has reached $18 billion on an annualized basis as of late 2025, nearly matching peer-to-peer stablecoin transfers at $19 billion. Monthly crypto card volume rose from roughly $100 million in early 2023 to more than $1.5 billion by late 2025 — a 106% compound annual growth rate.
Visa dominates, capturing more than 90% of on-chain card volume through early partnerships with crypto-native infrastructure providers. Across six Visa-partnered crypto card programs alone, total net spending rose from $14.6 million in January 2025 to $91.3 million by December — a 525% annual increase. Visa's stablecoin-linked card spending hit a $3.5 billion annualized run rate by Q4 2025, representing about 19% of total crypto card volume.
Mastercard is now aggressively counter-positioning. Beyond MetaMask, the company is hiring a Director of Crypto Flows to advance crypto-to-fiat infrastructure, weighing an investment in blockchain infrastructure company Zerohash, and rolling out its Crypto Credential verification service. As Mastercard CPO Jorn Lambert stated: "Blockchain capabilities and tokenization can reach their full potential, delivering true transformation."
The significance of these numbers becomes clear in context: the entire crypto card market was functionally nonexistent before 2022. It has gone from zero to $18 billion in three years. Self-custody cards — where users control keys until the moment of purchase — represent the newest and potentially most disruptive segment.
Most crypto cards on the market today are custodial. Coinbase Card, Bybit Card, and Crypto.com's Visa all require users to deposit funds onto the platform before spending. The exchange holds your crypto, converts it at the point of sale, and settles with the merchant. Functionally, this is no different from a traditional prepaid debit card — the crypto conversion happens behind the scenes, and the user has surrendered custody.
Self-custody cards invert this model. Assets stay in the user's own wallet — a MetaMask hot wallet, a Gnosis Safe smart-contract account, or a Phantom mobile wallet — until the exact moment a purchase is authorized. Only then does the card infrastructure draw the stablecoin balance, convert it to fiat, and settle through traditional card rails. The user never relinquishes their private keys to a third party.
This distinction matters for three reasons:
Counterparty risk elimination. Every custodial crypto card carries FTX-style risk. If the exchange fails, user funds are in the bankruptcy estate. Self-custody cards remove this vector entirely.
Composability. Self-custody wallets can simultaneously interact with DeFi protocols. A user can earn yield on Aave, hold the resulting aTokens, and spend from the same wallet — something impossible with custodial cards that silo funds in exchange accounts.
Regulatory clarity. The SEC's evolving framework increasingly distinguishes between custodial and non-custodial products. Self-custody card issuers avoid the regulatory overhead (and liability) of holding customer assets.
As of 2025, approximately 59% of crypto wallet users already use non-custodial wallets. Hardware wallet sales increased 31% last year. The market is moving toward self-custody by default — and cards are following.
On February 26, 2026, Consensys announced the general availability of the MetaMask Card across 49 U.S. states (excluding Vermont), including — notably — New York, which has historically been hostile to crypto-linked financial products.
The product stack:
Two tiers exist: | Feature | Standard | Metal ($199/year) | |---------|----------|-------------------| | Card type | Virtual | Physical metal | | Cashback | 1% in mUSD | 3% on first $10K/year | | Foreign TX fees | Standard | None | | ATM limits | Standard | Elevated | | Extras | — | Travel perks, exclusive events |
The card had previously been piloted in the UK and EU in 2024, then expanded to Argentina, Brazil, Canada, the EEA, Mexico, Switzerland, and the UK before the U.S. rollout. The nationwide U.S. launch represents the largest addressable market yet for a self-custody card product.
MetaMask Staff Product Manager Ale Machado framed the ambition clearly: success means U.S. users "spending from their self-custodial wallet without friction, without thinking about the crypto underneath."
MetaMask is not operating in a vacuum. The self-custody card segment is becoming crowded:
Gnosis Pay — The pioneer. A Visa debit card connected to a Safe smart-contract wallet on Gnosis Chain, offering up to 5% cashback through partner programs. Available across the EEA with planned expansion to the U.S., Mexico, Colombia, Australia, Singapore, Thailand, Japan, Indonesia, and India. Gnosis Safe wallets already custody over $70 billion in assets.
Phantom Cash — Solana's answer. A Visa card linked to the Phantom mobile wallet, with an integrated USD-backed stablecoin (CASH) issued through Stripe's Bridge platform. Currently in U.S. early access (excluding New York and Alaska). Users earn rewards on unspent stablecoin balances — effectively a self-custody savings account with a spending card attached.
Other contenders include 1inch's debit card (up to 2% rewards, but high fees erode the value), COCA's Visa card (tiered rewards up to 8% via token staking, but requiring lockups), and Ready's Metal Card (3% cashback in STRK tokens).
The competitive dynamics break down along three axes:
| Card | Network | Chain | Custody Model | Key Differentiator | |------|---------|-------|--------------|-------------------| | MetaMask | Mastercard | Ethereum/L2s | Self-custody | Largest wallet user base, Stripe/Bridge rewards | | Gnosis Pay | Visa | Gnosis Chain | Smart-contract wallet | Deepest DeFi composability | | Phantom Cash | Visa | Solana | Self-custody | Integrated stablecoin + savings yield | | Coinbase | Visa | Custodial | Custodial | Regulatory trust, U.S. market share | | Bybit | Visa | Custodial | Custodial | Up to 10% headline cashback (effective: 2-4%) |
The real story is not the cards themselves — it is the infrastructure stack enabling them. Three layers have converged:
Layer 1: Stablecoin issuance. Stripe's acquisition of Bridge for $1.1 billion in 2024 created the dominant stablecoin infrastructure platform. Bridge's transaction volume quadrupled in 2025. In February 2026, Bridge received conditional OCC approval to form a national trust bank, which would allow it to issue stablecoins, custody digital assets, and manage reserves under direct federal oversight. Bridge now powers both MetaMask's mUSD and Phantom's CASH — meaning Stripe sits beneath two of the three major self-custody card products.
Layer 2: Card program management. Monavate (formerly Baanx) provides the middleware that links blockchain wallet balances to card authorization flows. This backend synchronizes on-chain fund verification with traditional merchant settlement in real time. Without this layer, the self-custody model breaks — there is no way to authorize a card transaction against an on-chain balance without infrastructure that speaks both protocols.
Layer 3: Regulated banking. Cross River Bank (for MetaMask/Mastercard) and similar FDIC-insured institutions act as the issuing banks. This is non-negotiable: Visa and Mastercard will not process card transactions without a regulated issuing bank in the chain. The issuing bank underwrites the settlement risk — if a user's on-chain balance is insufficient, the bank absorbs the loss and pursues the user.
This three-layer stack — stablecoin issuance, card middleware, regulated banking — is the actual product. The wallets and cards are just interfaces.
From the webthreepedia economic-value-distribution framework, the self-custody card ecosystem creates a new value chain worth analyzing:
For every $100 spent on a self-custody card, approximate value distribution:
The critical insight: Visa and Mastercard capture the lion's share of value even in a "decentralized" self-custody model. The card networks' interchange infrastructure — negotiated decades ago with millions of merchants — is the irreplaceable moat. Self-custody changes who holds the assets before the transaction. It does not change who controls the payment rails after it.
This represents a fundamental tension in the self-custody card thesis. The products are marketed as DeFi meeting traditional finance. In practice, they are traditional finance co-opting DeFi wallets as customer-acquisition channels. Mastercard and Visa gain access to crypto-native users without building new rails. Wallet providers gain user stickiness and subscription revenue. The user gains convenience and retains custody. But the economic value still flows predominantly to the incumbent payment networks.
Crypto card spending has hit $18B annualized, with self-custody cards as the fastest-growing segment. Three major self-custody cards (MetaMask, Gnosis Pay, Phantom Cash) have launched or expanded in the last 90 days.
Stripe is the hidden infrastructure winner. Through Bridge, Stripe powers stablecoin issuance for both MetaMask's mUSD and Phantom's CASH. Bridge's volume quadrupled in 2025 and it is pursuing a national bank charter. Stripe may become the Visa of stablecoins.
Self-custody eliminates counterparty risk but not payment-rail dependency. Users keep their keys until purchase, but Visa and Mastercard still capture the majority of per-transaction economics. Decentralization stops at the point of sale.
The 59% non-custodial wallet majority creates a massive addressable market. Self-custody cards convert existing non-custodial wallet users into spenders without requiring them to change their custody preferences. MetaMask alone has over 30 million monthly active users.
Regulation is an accelerant, not a barrier. The SEC's increasing distinction between custodial and non-custodial products, Bridge's OCC charter approval, and FDIC-insured issuers (Cross River Bank) give self-custody cards a compliance advantage over custodial alternatives.
Self-custody crypto cards represent the most commercially viable bridge yet between decentralized finance and everyday spending. The MetaMask Card launch across 49 U.S. states, backed by Mastercard and FDIC-insured Cross River Bank, is a milestone — but the real story is the infrastructure maturation that made it possible.
Stripe's Bridge platform, with its quadrupled volume and pending national bank charter, has become the stablecoin issuance backbone for multiple major wallet providers. Monavate has solved the card-to-chain middleware problem. Regulated banking partners have accepted the settlement risk.
The economic reality, however, demands sobriety. These products do not disintermediate traditional payment networks — they extend them. Visa and Mastercard remain the value-capture layer. What has changed is the custody model: users no longer need to trust an exchange with their assets before spending them. In a market where the Fear & Greed Index sits at 11 — FTX-collapse territory — that counterparty risk elimination is not a feature. It is the product.
The question for 2026 is whether self-custody cards can scale beyond crypto-native users to mainstream consumers who do not yet have MetaMask installed. If they can, the $18 billion crypto card market is the floor, not the ceiling.