Cards You Can Load With Bitcoin

A bitcoin spend card is a conversion schedule wearing a payment card. The top-up rail gets the attention, but the decisive question is quieter: when does your bitcoin get sold? At load, the balance is dollars and the exposure ended before it began. At authorization, you ride the price to the terminal and take a disposal per purchase. Undocumented, you are guessing - and the majority of this market still makes you guess. That single field separates the operators who treat the bitcoin leg as a product feature from the ones who treat it as plumbing.

37 live cards take a bitcoin load today. 13 of them accept on-chain BTC directly; the rest convert in-app from an exchange balance. The fee surface is layered - rail, top-up, spread, subscription, ATM - and this page prices each layer, then shows where the models differ and where they quietly converge on the same trade: your coin for a claim.

The fee stack, layer by layer

Four layers sit between your wallet and the merchant, and operators disclose them with very different enthusiasm:

• The load rail. On-chain at 1 sat/vB base (2026-10-05 [5]) costs cents - the flat fee only punishes proportionally at small loads. In-app conversion skips the network entirely but moves you onto the operator's exchange pricing, where the spread lives.
• The top-up fee. Published and competitive at the strong operators: 0% at Bitpanda, Fold, Coinbase, Kast and most exchange cards; 0.9% at Bybit; 1% at AdvCash; 1.3% at Freedomia's USDT/USDC rail [4]; 3.5% at Goblin - the price of no-KYC intake. Median across the page: ~0%.
• The conversion layer. The one the "0% fees" headlines skip. Fold's 1% buy fee below its $5,000 Reserve threshold is documented [1]; the equivalent cost at operators that do not publish a spread is discoverable only by comparing the card's rate to a reference quote - the method is in the spread guide. 8 of the cards here publish a 0% FX fee; FX and spread are different layers, and 0% on one says nothing about the other.
• The subscription. Xapo prices the whole model as membership ($83.33/month, 0% FX [8]); Fold gates its 0% conversion and ATM waivers behind a $5,000 Reserve; KAST's tiers run to $10,000/year. The flat-fee shapes are converging on wealth products, not payment cards.

Live values from our database - details and caveats in each review. "-" = not published by the operator.

The sale-timing spectrum

Documented at-spend conversion is the market's emerging standard, and the wording is usefully consistent. Fold: the conversion to fund the card happens per its exchange terms, with a stated 1% fee below Reserve [1]. Blink: conversion "at the prevailing rate at the time of the Transaction" [4]. Coinbase's card sells at authorization. The pattern is deliberate - at-spend lets the operator market "spend your bitcoin" while keeping the conversion margin inside its own venue, and it gives the user continuous exposure, which some want and everyone should choose knowingly.

The undocumented remainder is the finding this page exists to keep score of. An operator that documents the conversion moment gives you a defined exposure window and a countable tax event; one that does not has made both ambiguous in its own favor. In our scoring the btc_flow dimension (25% of the total) exists to price exactly this - and "not documented" scores worse than any published answer.

What the strongest spend cards optimize for

The population splits into recognizable shapes once the fee stack is visible:

• Documented at-spend with rewards - Fold (0.25% BTC base, 1% conversion [1][2]), Krak (2% BTC back instant, Mastercard rate without markup), CoinZoom (0.5%, US). The pick for users who want BTC exposure to the terminal and a paper trail.
• Zero-FX global spenders - Bitpanda (0% FX, EUR rails, verified limits), Wirex (0% FX), Xapo (0% FX as a paid membership model [8]). The pick for multi-currency travel spend, not for BTC purists.
• Self-custody adjacent - ether.fi funds from a wallet you control (1% FX, 3% crypto cashback); Blink's balance can mirror a non-custodial wallet mode [4]. Structurally the honest versions of what most cards only imply.
• No-KYC entries - Goblin (3.5% top-up, 0% FX) and Freedomia ($5/month, documented limits) trade identity for fees; the no-KYC hub holds the full trade-off analysis.

Field notes

• Net the stack, not the headline. "0% fees" with an undisclosed spread loses to a 1% documented conversion the moment you compare a reference quote. One afternoon of testing a $50 load beats any fee table.
• Match the sale model to your view. Bullish on bitcoin? At-spend (Fold, Coinbase, Blink [4]) keeps exposure to the terminal. Want the card to be pure plumbing? At-load or a stablecoin rail (KAST [3]) removes the price risk entirely.
• Keep balances small by design. The balance is an unsecured claim from the moment of load. Two weeks of spending is a policy, not a paranoia - the structural alternative is the borrow lane, where the collateral legally stays yours.
• Automate the tax trail. At-spend cards generate a disposal per purchase [7]. Export monthly, not in April; the tax estimator counts the events before your accountant does.
• Re-read the terms quarterly. Fold replaced its entire rewards structure in September 2026 (Fold+ to Fold Reserve, 0.5% gated to 0.25% base [1]) - a documented change at a well-behaved operator. The pattern repeats market-wide, usually without the changelog entry.

Warnings and hard limits

• "0% fees" is a marketing tier, not a fact. Every zero on this page sits next to an unstated conversion spread or a paid tier that buys the waiver back. Price the stack end to end or the card is more expensive than it looks.
• The undocumented sale timing is not neutral. It transfers risk to you in the one moment (conversion) where the operator earns. Treat undocumented as a strike against the btc_flow score - which is exactly how our methodology weighs it.
• US issuers are not fungible globally. Fold and Cash App are US-only with full KYC; Freedomia's global no-KYC footprint comes with a sanctioned-country blocklist and $5,000/month ceilings [4]. The region columns in the table are the documented availability - the marketing map is not.
• Refunds land in the balance, not the chain. Returns on a converted balance arrive in fiat; on a Lightning-funded card like Moon's they land as on-chain BTC [9's sibling fact - see the Moon Bitcoin-Check]. Either way the refund path has its own fees and timing that the purchase flow never shows you.

The earn lane answers the other direction - getting bitcoin back on spending - and the Lightning hub covers the funding rail that makes small loads rational. Load small, know when the sale happens, and the spend lane does exactly what it says.

Sources

1. Fold terms and conditions - 0.25% base rate, 1% buy/sell fee, sale at authorization - accessed 2026-10-05
2. Fold cardholder agreement - prepaid fee schedule ($0 monthly, $3.50 ATM, $3.00 FX) - accessed 2026-10-05
3. KAST homepage - stablecoin-first funding with 2-5% auto-conversion on BTC/ETH/SOL - accessed 2026-10-05
4. Freedomia FAQ - 1.3% USDT/USDC top-up, 2.5% FX fee, USD balance model - accessed 2026-10-05
5. Blink cardholder agreement v2.3 - conversion "at the prevailing rate at the time of the Transaction" - accessed 2026-10-05
6. mempool.space fee API - 1 sat/vB across all confirmation tiers (2026-10-05) - accessed 2026-10-05
7. IRS FAQ on virtual currency transactions - exchange for other property is a taxable event - accessed 2026-10-05
8. Xapo Bank - direct BTC banking with no FX fees on card spending - accessed 2026-10-05