2026 virtual card provider comparison: operating status and fee structures

first look at who is still here, then talk about who is cheap

Search "which virtual card platform is the best" and most of what you find is rankings: whose rate is low, whose card opening is cheap, whose BINs are stable. These rankings were broadly useful in 2024, but they almost all assume one thing — the platforms on the list are still operating. That assumption no longer holds in 2026.

Spreading the timeline out makes it clearer: the four most-recommended platforms in the Chinese-speaking community — WildCard, DuPay, OneKey Card and Infini Card — all stopped service within 2025. Among them, the largest by volume was, before it shut down, the default answer in most "how to subscribe to ChatGPT" tutorials.

Timeline: how each of the four platforms exited

The four dates below are the skeleton of this wave of shutdowns. First, a note on the nature of the data: the platforms' original announcements may not all be accessible, and the dates and exit methods below were cross-checked from multiple public reports — they are a compilation of public information, not a record of dealings between us and these platforms.

The four exits differ quite a bit in form, but one thing is common: none of them went out through "gradual decline from mismanagement" — the entire business line was taken down, actively or passively. For users, that means a shutdown often gives you no preparation time.

This timeline is still playing out in 2026. In June of that year, Fiat24, a platform that marketed itself on a European license, also suspended new user registration from mainland China — contraction and shutdowns did not stop at 2025. Treating "will this platform be the next one" as a routine variable in platform selection, rather than as a low-probability event, matches the reality of these two years better.

Three structural reasons behind the collective exit

The four platforms come from different backgrounds: some were built by hardware-wallet teams, some by crypto-native teams, and some were payment tools aimed purely at domestic users. With such different origins all ending in the same outcome, the problem isn't individual teams' operating ability but the structure of the position itself.

1. The card isn't issued by themselves. The vast majority of these platforms are not card-organization members; they "rent" a bank's membership through a sponsorship model to issue cards. Once the upstream licensed institution tightens, gets penalized, or terminates the cooperation, the front-end card fails immediately — faster than the platform itself can prepare;
2. Compliance cost is rigid, and it only goes up. Cross-border fund flows, anti-money-laundering screening, identity verification — every item needs continuous investment; serving North American and European markets adds local registration and licensing requirements. This cost doesn't shrink just because the platform is small;
3. Thin revenue, squeezed from both ends. Raise rates and nobody uses it; lower them and compliance plus operations aren't covered, and the spread in between relies on user fund float — which is exactly the part regulators are most sensitive about.

Stack these three together and you get a position that's neither here nor there: neither the cost-spreading ability of a licensed bank, nor the cross-subsidization headroom of a large platform with other businesses. The industry has many postmortems of this path; for background, see this industry review of the retreat of the U-card model (external source, for verification only, not an endorsement). In our own four shutdown postmortems we only compile facts, and we don't make predictions.

Understanding this matters because it directly changes the structure of the answer to "which is better": when the leaders of an industry exit en masse, the remaining choice is no longer "pick one of these platforms" but "which kind of model can survive longer".

Fee structure comparison: five platforms on one table

The table below puts us and the four platforms together. First the basis, then the numbers — the point of this table isn't "whose number is smaller", but which step collects the fee and what conditions are attached.

Table note: the numbers in the right four columns come from third-party public compilations before these platforms shut down; many of those pages are promotional, sources conflict with each other, and the numbers are no longer valid after service stopped — they're listed only to show "what the price structure of the time looked like" and do not constitute a quote for any platform. The PIKABAO column follows the on-site pricing page.

Reading this table through shows two things:

• A low rate is usually made up elsewhere. DuPay's lowest rate is tied to the highest card-opening fee and card tier — a "pay the fixed cost first, enjoy the low rate later" model; Infini charges nothing for top-ups, but it's a stablecoin product aimed at corporate settlements, with barriers and usage entirely unlike regular subscription users;
• Fixed outlays are easier to overlook than rates. Monthly fees and validity renewals in the table weigh far more on light users than the rate gap — a card with a 2 USDT monthly fee costs 24 USDT a year, the equivalent of a full year of top-up fees for someone loading 20 USDT a month.

That's also why we put our column on the far left: a comparison table should be compared on today's terms, not propped up against others' historical prices. The four platforms' numbers stay in the table to help read the structure, not to rank anyone.

Total-cost example: for the same usage, where the money goes

Turning rates into money makes comparison easier. Below we calculate two usage bands: a light user loading 20 USDT a month onto the card (240 a year), and a medium user loading 100 USDT a month (1200 a year). The basis is uniform: "fixed outlay of opening the card once + one year of holding cost + one year of top-up cost", excluding surcharges on non-USD spending.

Example basis: calculated at the publicly posted rate tiers before each platform shut down; the four platforms' numbers are already invalid and only serve structural comparison. PIKABAO follows the on-site pricing page. Transaction surcharges such as "non-USD spending" aren't included.

The conclusion this table supports needs to be stated plainly, without fudging:

• In the light-usage band (20 USDT a month), our total cost is the lowest. The reason is that at this usage level, fixed outlays like monthly and annual fees account for the bulk — and we're at 0 on both;
• In the medium-usage band, on pure numbers we are not the lowest. Platforms modeled on "low rate + free top-ups" come out cheaper. We're writing this out directly — because a table that only compares the band where you win has no reference value;
• And those cheaper numbers belong to platforms that have all shut down. That's not a coincidence: compressing rates to that level is itself one of the reasons they couldn't sustain operations.

Where "cheaper" comes from: no charge in three places, one charge in one

Break the fee structure open and what we do is actually simple — three steps charge nothing, one step charges, and it's a single percentage.

This design has a direct point of comparison: tiered rates look like they give you a "cheaper choice", but they front-load the fixed cost — to get a 1.0% rate you first pay a 50 USDT card-opening fee or reach a tier. People who don't use that much volume end up paying for a rate they never use.

The full fee table and worked examples are listed item by item on the pricing page; the complete table of all four fee blocks is also covered in What is the total cost of using a card.

Four common misconceptions about "which is the best"

Summarizing the public discussion in this wave of shutdowns, four claims come up the most often, and every one needs correcting:

What choices remain in 2026

After crossing the four platforms off the list, three realistic paths remain. Between them it isn't "which is better" but "which kind of risk you're taking on":

The third type is exactly where the shut-down platforms turned — no more cards, top-up only. It's enough for someone subscribing to just ChatGPT, but for an ad buyer it's no substitute at all.

For the vast majority of individual users and small and mid-size buyers, the realistic option is the first type. Its selection criteria don't need to be complex: can the entity information be found publicly, is the exit rule written before you put money in, and can the upstream and BINs be explained. All three can be verified before opening a card, without relying on any verbal promise.

Our situation, laid out against these three criteria for you to check: the operating entity, registration number and filing information are public; the balance-withdrawal rule (free, 48 hours after the card's last transaction) and the refund handling fee (flat $3) are written on the pricing page, not something you only learn at exit; card types, BINs and issuing regions are shown in real time on the products page. These three can't prove we'll never change, but they guarantee one thing — if you need to leave, the rules were already written down before you came in.

FAQ

Further reading

a side-by-side comparison of the four shut-down platforms, focusing on the timeline and fee structure. The order for selecting a platform (define needs first, then screen with one-vote-veto items) is covered in How to choose a virtual card platform, not repeated here; the four platforms' balance handling and migration checklists are in WildCard, DuPay, OneKey Card and Infini Card.

For the concept and the funding flow, see U cards: definition and differences from virtual credit cards; the complete table of all four fee blocks is in What is the total cost of using a card; how to move card money back to the account is in Can card balance be moved back to the account; how to pick BINs and regions is on the products page.

Official sources: the mechanism of the prepaid card category follows the Visa official prepaid card page (external source, for verification only, not an endorsement); the regulatory stance that prepaid products should disclose fees in advance references the CFPB prepaid card consumer page (external source, for verification only, not an endorsement); the industry-model analysis cites the industry review of the U-card retreat (external source, for verification only, not an endorsement). External links are for verification only and do not constitute an endorsement; some links may not open directly in certain network environments. The dynamics of the four platforms involved in the example are compiled from public information only; their arrangements and policies may continue to change, so defer to the latest official announcements.