Crypto card referral programs

What the programs actually pay - verified, not marketed

61 programs on this page carry referral or affiliate mechanics, and the verification picture splits cleanly: 64% have independently verified fee data on the cards themselves, while the referral programs divide into fully published tables (one), fixed bonuses (three), and apply-to-learn programs (the rest). The one program that publishes everything is the instructive one - because its numbers show how small card-spend economics actually are next to trading-fee economics.

The three machines

The distinction between the machines is the behavior they pay on, and the behavior determines the durability - trading fees recur as long as the user trades, card spending recurs as long as the user spends, bonuses pay once. Trading-fee revshare is where the money is: KuCoin up to 60% of referred trading fees on permanent relationships [3], OKX up to 50% in USDT [4], Coinbase Advanced 30% in USDC under its published agreement [10]. Card-spend share is the rare structure: ether.fi's table pays 0.1-0.3% of referred card spending plus 20-25% of card fees across four tiers, lifetime, uncapped, paid in USDC [1]. Consumer bonuses are the third machine: Gemini's $50-plus-$250-per-five card referral [5], Nexo's balance-linked up to $2,500 [2], Crypto.com's milestone program with a $10,000 lifetime cap [7]. The structures are not interchangeable - they pay on different behavior, which is why the qualifying conditions matter more than the headline.

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

What the fine print actually constrains

The published agreements share three rules, and they define what a promoter may do - worth reading before promoting anything, because the consequences (fee forfeiture, program exclusion) attach to the promoter, not the program. No trademark bidding on paid search - Coinbase's agreement threatens forfeiture of at least 30 days of fees [10]. No return promises - ether.fi's own table carries an "estimates" label with no assurance attached [1], and Coinbase bans implied representations about investment outcomes [10]. And rates change unilaterally - RedotPay's archived terms cap recourse at the rebate balance [9]. The binding mechanics also differ from legacy affiliate marketing: relationships attach at registration and persist for the account's life [3][4], which makes the first-touch discipline (clean links, documented chains) the referral equivalent of the first-load discipline.

Warning: Referral rewards are program terms, not entitlements - the same one-way repricing that governs card rewards governs referral rates, with less public scrutiny. Payout timing matters too: ether.fi pays on the 15th of the following month [1], RedotPay accrues daily and pays within two business days [9], and Coinbase's simple referral pays in 5-15 days [6]. A rate you cannot verify against a payout history is a rate you do not have.

The card-spend economics, honestly

The only fully public card-spend structure - ether.fi's - pays $10-25 per new member depending on tier, 0.1-0.3% of their ongoing card spending, and 20-25% of card fees [1]. Run the arithmetic: a referral spending $500/month generates $0.50-1.50/month of revshare. That is the honest size of card-referral income per user - meaningful at scale, trivial at friendship scale, and structurally different from the trading-fee programs where the same referred user generating exchange fees produces 30-60% of those fees [3][4]. The consumer bonuses are larger per event ($50-2,500) but gate on the deposit conditions discussed above; the balance-linked Nexo structure in particular transfers the market risk to the referrer, since the payout depends on the referred user's holdings holding value. 8% is the best card reward rate in this population - the referral layer pays a fraction of what the card itself pays the user, which is the correct order of things.

Taxes and disclosure

Referral income is not card cashback: it is compensation, taxed as income at fair market value in most jurisdictions, with a disposal event when the tokens are sold. The US rebate rule (basis reduction, not income) does not apply - that treatment belongs to purchase rewards, and conflating the two is a filing error. The distinction matters at filing: cashback attaches to your spending, referral income attaches to someone else's signup. And disclosure is the other half of the compliance frame: most programs' terms require promoters to disclose the relationship, and the no-promise rule [10] covers the claims you make about the card, not just the program. If you recommend cards publicly, the cashback hub is the companion page - its cap and token-risk analysis is what honest promotion actually cites, and the fees guide is the reference behind every honest "the card is good, the rate is fair" claim.

Reading this page

The one structural takeaway before the list: card referral programs are downstream of the cards themselves - a rate on a card nobody should use is a rate worth nothing. The ranking orders the 61 programs by card quality - deliberately, because a referral program attached to a weak card is a bad recommendation regardless of its rate. Verify rates against live program pages (the archive-sourced RedotPay number [9] is explicitly dated), document your referral chain at signup, and treat the apply-to-learn programs as pricing conversations rather than published offers. The no-KYC hub and the credit hub round out the picture for the two audiences most likely to be recommending cards: privacy-conscious users and rewards maximizers. The card pages carry each program's affiliate fields with check dates - that is where the numbers above live when they change.