Give every AI tool its own card number, with a spending limit you control, and a surprise charge stops being a surprise. That is the idea behind the Foton Card Virtual Card, sold at fotoncard.com, and it is the part of Hello FotonCard's four-minute video I would keep even if you never open an account: the trick works on any card that lets you set a limit.
The video was published on July 8, 2025, so the sign-up process and the user figures it shows are the ones the company was advertising then.
What the Foton Card virtual card is
A virtual card is a set of card details (a number, an expiry date and a security code) that exists only online. Merchants run it through a card network the same way they run a plastic card, so it works at any checkout that accepts that network. There is nothing to post, and you can create one per merchant, cap it and delete it when you are done.
Foton Card comes across as a payment service for online spending in general, not only software trials. Its marketing shows the address fotoncard.com, the lines "Trusted Payment Service" and "Secure Pay, Global Way", and logos from AliExpress, Trip.com, PayPal, Google, Facebook and AWS as the kinds of merchant the card is built for. I read those logos as marketing. Whether a card works at a given checkout is decided by that merchant and the card network, and the answer can change without notice.
The problem the creator describes will be familiar to anyone who has signed up for a free trial. The tools ask for card details before you see the product, and the card stays on file afterwards. They list ElevenLabs, RunwayML, Phind, Jasper, Kaiber, Pika, Synthesia, Descript, Opus Clip, ChatGPT and Claude as tools they had tried in the previous two months, and say each one either charges monthly or asks for a card upfront.
A card number you can throw away is a cheaper way to find out whether a tool is worth paying for.
One card per AI tool, with a limit you set
The method takes four steps.
- Create a new card in your Foton Card account for the tool you want to try.
- Set that card's spending limit to $0 or to just enough to cover the trial charge.
- Enter the card details at the tool's checkout instead of your everyday card.
- If you decide to keep the tool, raise the limit on the same card; if you do not, leave the limit where it is and delete the card.
The $0 limit is the part worth understanding. A payment above the limit fails at authorisation, so the merchant never gets the money, and you have not handed over a card number tied to your main account. "If I like the tool, I update the limit and continue. If I don't, the card blocks the charge and I delete it."
The failure the method targets is the one the creator ran into: "I forgot to cancel a few, one charged me $29 without warning, one even renewed for a yearly plan". A trial that turns into a monthly charge is annoying. One that turns into an annual plan is a bigger bill, and it usually arrives because nobody read the terms.
I think the method is sound as far as it goes, and the limit does less work than it looks. A limit set to cover the trial still allows any charge up to that amount, so a merchant that bills more than you expected, or bills in another currency after conversion, can land inside the cap. The card also does not end the subscription. A blocked payment is a failed payment, not a cancellation, and plenty of services keep the account open, retry the charge and email you about it. If you want the subscription gone, cancel it in the merchant's own settings as well, then delete the card.
There is a second limit worth knowing about. Some merchants check that a card is valid by placing a small authorisation hold before the first charge, and a card with a $0 limit can fail that check, so you may not reach the trial at all. The creator's workaround, setting the limit to just enough, handles this. Zero does not.
Signing up without KYC
KYC means know your customer: the identity checks a financial service runs before it lets you move money, usually an ID document, a selfie and sometimes proof of address. The creator says Foton Card does not require it, which is how they signed up instantly and created a first card in minutes.
For a reader, the trade-off runs both ways. Less paperwork means less personal data sitting with a card provider, and no wait for a document review before you test a tool. It also means the provider may know very little about who is spending, which tends to make disputes slower and account recovery harder, since there is less on file to prove an account is yours.
Card programmes change, too. Identity requirements are set by the companies that issue and sponsor cards, and they can tighten when rules or banking partners change, so a sign-up that needs nothing this year can ask for documents the next. If you plan to keep a balance on a card, check the current requirements on the site rather than trusting a video from 2025.
The numbers Foton Card shows on screen
Four figures appear as a summary of the business.
| Metric | Figure shown | | Users | 100K+ | | Countries | 200+ | | Volume processed | $1B+ | | Years running | 5 |
"100K+ users" counts accounts, not active spenders, and "$1B+ volume" is the total value processed, which says something about scale and nothing about whether any individual cardholder was treated well. "5 years running" dates the service to around 2020 if the video is from mid-2025, which is a reasonable stretch for a card programme and still short enough that one bad year could undo it.
"200+ countries" is the claim I would test first, because it decides whether the product is available to you at all. Card programmes rarely treat every market the same way: availability depends on the issuer's licences and on local rules for the currency the card is denominated in, so check the list against your own address before you sign up.
All four figures are the company's own, shown in the company's own video. Treat them as marketing until a third party confirms them, the same way you would treat any user count on a landing page.
Top-ups, custody and the fees to check
Crypto cards generally work in one of two ways, and the difference matters more than the brand on the plastic.
In the custodial model, you send crypto (often a stablecoin such as USDT) to the provider, it converts the funds to a fiat balance, and the card spends from that balance. The provider holds the money, which makes your balance a claim on a company: if the programme freezes accounts, changes its banking partner or runs into trouble, your ability to spend depends on their support desk and their solvency.
In the self-custody model, the funds stay in a wallet you control and the card authorises payments from it, converting at the moment of payment. You keep the keys, so the counterparty risk is smaller, but the risk moves to the wallet and to whatever contract or approval lets the card spend. A wrong approval or a lost key is your loss, not the provider's.
The fee side has a standard shape across the industry, and each line changes the maths of a small trial:
- An issuing fee for each new card, which matters if you create one per tool.
- A monthly or annual fee for the account, which matters if you keep it open between trials.
- A top-up fee, flat or a percentage, which matters most on small amounts.
- A conversion spread when the card's currency differs from the merchant's, which is easy to miss because it sits inside the rate instead of on its own line.
- ATM and decline fees, which matter if you take cash out or if failed charges are billed.
The way to judge them is to price your smallest realistic purchase, not your largest. A fixed monthly fee is a large share of a small trial charge, and a percentage top-up fee hurts more when you fund $20 at a time than when you fund $500. If the card advertises rewards, read the conditions before counting them: cashback on cards in this category often comes with a token you have to lock up, a staking tier that sets the rate, or a monthly cap, and a reward that requires you to hold a volatile token is a position rather than a discount.
The video does not show a top-up, a conversion, an ATM withdrawal or any fee schedule, and it does not name the issuer, the card network or where the balance sits, so read the provider's own terms and fee page before you fund a card.
Spending crypto is usually a taxable sale
Paying with crypto is not like paying from a bank balance in most tax systems. In many countries, handing over coins or tokens for goods or services counts as disposing of an asset, and capital gains rules apply to the difference between what the asset cost you and what it was worth on the day you spent it. The subscription can be small and the tax question still applies.
Stablecoins are not automatically exempt. They are treated as crypto assets in many jurisdictions, so a USDT top-up and a later payment can both create a record you are expected to keep, even if the value barely moves.
This is where a card statement stops being enough on its own. A statement usually shows the fiat amount the merchant received. It does not show what you originally paid for the coins, which is the number the gain is calculated from. I would keep a simple record of each top-up and each payment: the date, the amount, the coin and the value in your local currency at the time. Rules differ from country to country and change, so check how your own jurisdiction treats card spending before you build a habit around it.
The risks that matter more than a surprise charge
Control is the product here, and it is real: a limit and a delete button do more than a reminder email. The risks sit around it.
The first is that the company is marking its own homework. The channel is called Hello FotonCard, the video ends with an invitation code, and the case for the card is made by the people selling it. That does not make the method wrong, and one card per merchant stands on its own. It does mean the smooth sign-up, the four figures and the no-KYC line are the company's account of itself, and I would want a second source before treating any of them as settled.
Then there is the balance. Money you have topped up and not spent sits somewhere, and in a custodial setup it sits with the provider. Keep only what you plan to spend in the next few weeks, and treat a large balance on a card as a loan to a company you have not checked.
The gap between blocking and cancelling is the risk the method leans on. A merchant that cannot charge you may suspend the account, keep the subscription live, or send the debt to a collection process if the terms allow it. Deleting a card ends your ability to pay. It does not end your agreement.
Access comes next. A card that fails at authorisation can fail at the wrong moment, and a virtual card can be declined where a merchant wants extra verification for a subscription. If a tool matters to your work, a throwaway card is the wrong place to save.
Privacy is the last one. A virtual card hides your main card number from the merchant. It does not hide you: the merchant still has your account, your email and your usage, and the card provider still has the transaction record. A card number you can throw away is not an anonymous identity.
What I would check before trusting a virtual card
- Confirm the site address before you type a password. The video gives fotoncard.com, and lookalike domains are the standard way accounts get stolen.
- Read the fee page and price your smallest planned purchase, including the conversion spread.
- Check that your country is supported and what verification the provider asks for today, not what a 2025 video said.
- Fund a small amount first and make one purchase before you rely on the card for anything recurring.
- Keep idle cards at a zero limit, and delete cards you no longer use.
- Cancel subscriptions in the merchant's own settings, then remove the card.
- Keep a record of every top-up and payment for tax.
The habit worth keeping is the one the creator demonstrates, not the brand: one card number per service, a limit that matches what you agreed to pay, and a delete button you actually use. That works on a Foton Card account and on any other card that lets you set limits.
If I were starting, I would skip the $0 card for the first attempt and set the limit just above the trial amount, because the zero-limit card is the one most likely to fail a merchant's verification check.
