Telegram Stars vs cards: which rail for which product (2026)

TL;DR. You don’t choose the rail — what you sell chooses it for you. Digital goods, channel subscriptions, paywall posts and donations are Stars-only. Physical goods and real-world services go through a card provider. Telegram’s developer terms make this binary, so no product legitimately offers both buttons. The decision that’s actually yours is what you sell — and that decision hands you a fee structure, a payout cadence and an audience along with the rail.

Telegram now ships two real payment rails. Stars is native, one-tap, no card form, instant. Cards are via a fiat payment provider: checkout still happens inside Telegram, and the money lands in your bank account in your provider’s currency — eighteen are supported, from USD, EUR and GBP through to BRL, INR and IDR.

Most write-ups frame this as a choice you make. It isn’t. Telegram assigns the rail from the product category, and the interesting question is what that assignment costs you. Here’s the practical breakdown.

What each rail actually is

Stars is Telegram’s built-in micro-currency. Customers buy Stars in-app once (in their app store), then spend them across any bot, channel or paid post on Telegram. From your side, you set a Stars price on the product, Telegram debits the customer’s balance, your bot’s Stars balance goes up. You withdraw later per Telegram’s payout policy.

Stars work for: digital downloads, channel subscriptions, paywall posts, donations, and services you deliver digitally.

Important, and it runs both ways. Telegram’s terms for bot developers state that Telegram doesn’t process payments for physical goods and services at all — those must go through a third-party provider — while transactions for digital goods and services must be executed exclusively through Stars. So there is no card button to toggle on a paywall post, and no Stars button on a parcel you ship. Selling a digital product through a card provider is the case Telegram explicitly warns about: it can send a notice to the account that owns your bot.

Cards in Telegram go through a payment provider linked to your bot. The customer sees a Telegram-native checkout sheet, taps Pay, enters card details once (saved for next time), and your provider routes the funds to your bank account in your country’s currency. Pretty much identical experience to Stars, except the money never converts through Stars.

The real economics, without the fairy tales

Both rails take fees. Anyone telling you the math is simple is selling something.

Stars side. Telegram takes its cut. The cut depends on the platform the customer used to buy Stars in the first place — Apple and Google each take their own slice when Stars are purchased through the iOS or Android app store, and Telegram takes another slice on top. The exact percentage varies and Telegram tweaks it. Your effective revenue per Star sold is lower than the gross sticker price — model it as “I keep roughly 70% of the sticker price” and you’ll be in the right ballpark for digital goods. On desktop or web app, the Apple/Google store cut disappears, so your share is closer to the top of the range — but Telegram’s own fee still applies.

Card side. The fiat payment provider takes ~2–4% per transaction, plus a fixed fee for some methods. There’s also a settlement delay (T+1 to T+7 depending on country and provider), and you handle fiat invoicing and VAT in your jurisdiction yourself.

AdminHub’s cut on either rail: 0%. We monetize on Pro subscriptions to creators, not on customer transactions. So the only fees that touch your revenue are Telegram’s (on Stars) or your card processor’s (on fiat).

0% is our cut, not the whole cost. Telegram’s own Star economics sit underneath it: buyers pay roughly $0.019 per Star, and Telegram’s developer terms put the payout at $0.013 — so plan revenue on the second number, not the first. See what withdrawing Stars actually pays.

What the Stars rail is good at

Read these as reasons to shape an offer as a digital product, not as a rail you can bolt onto something physical.

• Impulse-buy digital goods under ~$20. One tap, no card friction. Customers who’d abandon a card form will complete a Stars purchase.
• Channel subscriptions. Telegram Stars subscriptions (launched 2025) is the native rail for monthly recurring access. Auto-renewal works without you wiring up a separate billing system. Cancel and manage live in the Telegram UI, not yours — fewer support tickets.
• Paywall posts. One-time unlock of a single post (newsletter, exclusive, premium piece). The whole flow is in the chat — no detour to an external web checkout.
• Donations. Free amount, instant, no friction. Carving out a custom donation form is heavier than the donation itself.
• Audience under ~30 years old. Stars-native generations don’t blink at the one-tap experience. Card forms feel like 2018 — use this as a rough guide, not a law; test it against your own audience.
• Cross-border audience. Customers from many different countries pay without you setting up local card processing — Stars is a single currency, and conversion happens on Telegram’s side.
• Anything where speed of conversion matters more than fiat reporting. Stars is the win.

What the card rail is good at

These come with selling something physical or delivered in the real world — that’s what puts you on this rail.

• Single purchases over ~$50. Buyers want a receipt, a card statement, a refund process they recognize — and above that ticket size the familiarity converts.
• B2B and invoiced sales. Companies don’t put Stars on expense reports. They want a card charge with an invoice attached.
• Regions where Stars adoption is thin. In some countries the Stars purchasing flow is still unfamiliar, and a card checkout converts better. If the same value can be packaged either way, those geos are an argument for the physical format — not for bolting a second button onto a digital one.
• Recurring fiat for accounting compliance. If you need monthly invoices with VAT for your accountant or your customer’s accountant, cards are simpler.
• Audience over ~40 years old. Card experience is what they typically expect. As a rough guide — not a rule — don’t force a new payment habit on someone you’re just trying to sell to.
• Anything where you need the money in your bank account by Friday. Card settlement is more predictable than the Stars payout cadence.

What’s new in 2025–2026: Stars subscriptions and paid media

Telegram opened native Stars subscriptions in 2025 — recurring 30-day billing for channels, with auto-renewal handled by Telegram. Before that, monthly access to a private channel meant duct-taped solutions: invite-link bots, external billing portals, manual revokes when payment failed. Now it’s one Telegram-native primitive.

This is the rail to use for paid channels, creator subscription models, premium newsletter access, and gated communities. Set the price, set the period, ship.

Paid media unlocks (paywall posts) and donations followed — same theme: native Telegram primitive instead of an external workaround. If you’re still routing channel subs through external links and revoking access on scheduled jobs, you can rip that out.

You can’t run both rails — so run two products

The advice you’ll read elsewhere is to show both buttons and let the customer pick. Under Telegram’s terms that isn’t available: each product category is assigned one rail, and a checkout offering both on the same item is offering one of them against the rules.

What you can do is sell on both rails through different products. A course sold as digital access takes Stars; the printed workbook that goes with it ships and takes a card. A consultation delivered over a call takes Stars; the same consultant’s on-site visit takes a card. Same catalogue, same storefront, two rails — assigned honestly rather than toggled.

The classification is the seller’s call, and it’s worth making carefully. “Digital” means the buyer receives it inside Telegram or as a download; “physical or real-world” means something ships or happens in person. A workshop recording is digital. The same workshop attended in a room is not.

In AdminHub the storefront applies this for you: mark a product as physical and it takes card payments, and the shipping fields appear alongside. See it on storefronts here.

The decision rubric

Start from what you’re selling — the rail follows:

1. Digital goods, downloads, paid access? Stars. No provider to set up, so you can ship today.
2. Paid channel or paywall posts? Stars subscriptions and paywall posts — don’t bolt on a separate billing system.
3. Services? Whichever matches delivery: over a call or in-chat is digital and takes Stars; on-site takes a card.
4. Physical goods? A card provider, which means connecting one before your first sale — budget for the approval time.
5. Want both rails? Then you need products in both categories, not two buttons on one product.

The trap worth naming: a physical product can’t be launched as fast as a digital one, because the provider approval sits between you and your first sale. If you’re testing whether an audience will pay you at all, test it with something digital — that answer arrives in a day rather than a fortnight, and it costs nothing to find out.

---

For the full playbook on launching a Telegram storefront — bot setup, product types, customer support — see How to sell on Telegram in 2026.