Comparison

Irreversible settlement against universal reach

TON settles peer to peer with no intermediary, no merchant account and no chargebacks, and is final once confirmed. Cards reach anyone, settle to a bank, and can be reversed weeks later. TON suits a crypto-native audience and a low-refund product; cards suit everything else.

Irreversible settlement against universal reach: the decision in facts

Compared against: Card payments
Dimensions compared: 8, including reversibility and intermediary
Where we lose: Reach: Nearly everyone has a card; a small minority holds TON

As of 2025-10-01, Telegram Bot API 13.4

What is actually being compared

The structural difference is who can undo the payment. A card transaction can be reversed by the issuer weeks after the fact, which is protection the customer has paid for through the fee. A TON transaction cannot be reversed by anyone once confirmed. That single property drives nearly every other difference between them, including the fee.

For a business, irreversibility is genuinely attractive where the product is delivered instantly and disputes are rare — digital goods to a crypto-native audience, for instance. There is no chargeback exposure, no dispute evidence to compile, and no reserve held against future reversals. Businesses that have fought a chargeback know what that is worth.

It is a liability where refunds are ordinary. A restaurant, a rental company or a clinic issues refunds routinely, and on a chain each one is a new outbound payment somebody has to execute. The absence of a chargeback mechanism also means the customer has no recourse if you fail to deliver, which sophisticated buyers understand and general consumers do not — and that asymmetry is itself a reason a general audience does not adopt it.

Side by side

Dimension	card payments	A custom build
Reversibility	Chargebacks up to months later. Customer protection you pay for.	None. Final on confirmation, with no arbiter.
Intermediary	Acquirer, scheme and issuer, each taking a share and imposing rules.	None. The network settles between two wallets.
Merchant account	Required, with underwriting that excludes several categories.	Not applicable. A wallet address is all that is needed.
Reach	Anyone with a card, which is nearly everyone.	People holding TON or a jetton, which is a specific audience.
Price stability	The amount quoted is the amount paid.	Native TON moves between quote and payment. A stablecoin jetton removes this.
Settlement speed	Authorised instantly, settled to your bank in days.	Confirmed in seconds and available immediately.
Fees	Percentage plus fixed fee, plus any reserve your acquirer holds.	Network fees only, and small.
Regulatory position	Well established, with clear obligations.	Accepting virtual assets may be regulated. In Dubai, VARA is the relevant authority.

Which one fits your situation

Choose card payments when

Your customers are a general consumer audience rather than a crypto-native one.
Refunds are ordinary in your business, which makes irreversibility a recurring manual cost.
You need funds settling into a bank account for ordinary business operations.
Your customers expect the protection a chargeback provides, and would notice its absence.
You can get a merchant account for your category and settlement in your market.

Choose a custom build when

Your audience already holds TON or USDT and finds a wallet payment easier than a card.
Your product is delivered instantly and disputes are genuinely rare.
No card processor will underwrite your category, which for virtual asset businesses is usual.
Chargeback exposure has become a real cost you would rather remove entirely.
You are selling across borders where card acceptance is inconsistent and expensive.

Where we lose

Reach, decisively. Nearly everyone has a card and a small minority holds TON, which makes this a rail for a specific audience rather than a general one.
Refund handling. Every refund is a new outbound payment executed by a person, which is fine occasionally and a job if it is routine.
Customer protection. There is no recourse if you fail to deliver, and that asymmetry is a legitimate reason for a general consumer to decline.
Price stability with native TON. The amount moves between quote and payment, which makes purchases feel like a bet — a stablecoin jetton fixes it and adds its own fee complexity.

If you are already on it, how the move works

Establish what share of your audience actually holds TON
Offer it alongside cards and measure uptake rather than assuming. For a crypto project it can be most of the audience; for a consumer business it is usually a rounding error, and that decides whether any of this is worth building.
payments-api
Quote in a stablecoin rather than in native TON
A price that means the same thing when the customer presses send converts materially better. Quoting in a volatile asset makes a purchase feel like a bet on timing.
commands
Verify against the chain, never against the client
The backend records what it expects and watches for it at a confirmation depth scaled to the amount. Fulfilling on a wallet's success callback is the mechanism behind most losses on this rail.
webhook
Write the refund policy before launch, not after the first request
A refund is an outbound payment somebody executes manually. Deciding who, within what window, and how it is recorded is a business process that has to exist before it is needed.
commands
Keep cards for everyone else
This is a second rail rather than a replacement in almost every case. Removing cards to force adoption removes customers, and the ones removed are the majority.
payments-api

Questions people ask when deciding

Is accepting TON regulated in Dubai?

It may be, depending on what you are selling and how you handle the funds. VARA regulates virtual asset activity in Dubai, and the question belongs with your counsel before the build rather than after. The protocol being permissionless does not make the activity unregulated.

What happens when we need to refund someone?

You send a new payment. The original is irreversible and there is no arbiter, so refunds are a business process rather than an API call. If refunds are routine in your business, that is a strong argument against this rail.

Should we quote in TON or in a stablecoin?

A stablecoin, in almost every commerce case. Native TON moves between the quote and the payment, which makes the customer feel they are gambling on timing, and abandonment rises with visible movement.

Can we drop cards entirely?

Only if your entire audience is crypto-native. For anyone else, removing cards removes the majority of customers, and the ones you keep were never the constraint on growth.

Is it really cheaper than cards?

On fees, substantially — network fees are small and there is no acquirer share or reserve. Whether it is cheaper overall depends on the manual refund handling and the customers you do not reach, both of which are real costs that do not appear on a fee schedule.

When can we treat a TON payment as complete?

At a confirmation depth scaled to the amount, verified against the chain. Not when the wallet says it sent it — a wallet is software the customer controls, and its report is a claim rather than evidence.