All systems operational 12 elite locations · Europe & Asia Pay with crypto · 0 KYC
Home / Guides / Buy a VPS With GRAM (TON) — No KYC
Guide

Buy a VPS with GRAM (TON) — no KYC payment walkthrough

Buy a VPS with GRAM (TON) on VPSbit — no KYC

Pay a VPS invoice with GRAM (TON): pick the coin, send native TON to the UQ address, credit lands in 1–3 minutes. Core from $4.80/mo annual-eff, no KYC.

Short answer

Pay a VPS invoice with GRAM (TON): pick the coin, send native TON to the UQ address, credit lands in 1–3 minutes. Core from $4.80/mo annual-eff, no KYC.

The invoice wants the chain asset, not a Telegram token

The label on the checkout says Gram (TON), and that label is the most misleading word on the page — usefully misleading, once you know why. What the invoice actually accepts is native TON, the live asset of The Open Network, not the cancelled 2018 Telegram token and not any ERC-20 contract that borrowed the name. If you send the chain asset once, to the printed address, at the printed size, the credit is mechanical.

The mechanics behind that sentence are plain. At checkout you pick Gram (TON) from nineteen coins, the page prints an exact figure converted at the live rate, and under it sits a single-use address starting with UQ. That prefix marks a user account on TON — the address formats are documented in the official TON documentation — and a watcher reads inbound transfers against the printed amount. No memo, no reference field, no identity data: the pair of amount and address is the entire message.

Crediting is a match, not a guess. The toncenter watcher — the same API service that powers public explorers, described at toncenter.com/api/v2 — looks for one inbound transfer of the exact printed figure to that one address, and both sides of the pair have to line up. The invoice stores no tolerance for near-enough payments. Numbers worth keeping: the catalog carries 19 coins across 12 locations, and GRAM (TON) is one entry among them, not a special rail with its own rulebook.

Where it breaks is asset confusion, and it breaks loudly. A jetton — a token issued on top of TON — can present the same address screen in some wallets while moving a completely different asset underneath. Any ERC-20 named GRAM on Ethereum is a stranger twice over. The chain itself, defined in the TON whitepaper, has one native asset; a wallet screen labeled Jetton is, by construction, offering something else. A wrong-asset send does not credit, and recovery runs through a ticket rather than a resend.

So the working mental model is the one you would apply to Bitcoin versus a wrapped token on some other network: the invoice wants the native coin, sent once, at the printed size, from a wallet whose keys you hold. Every other question on this page — timing, wallet behavior, troubleshooting, dedicated metal — follows from that single constraint, and the rest of the guide just walks its consequences in order.

VPS from $4.80* — Core on the annual term (list $6/mo), one invoice for 12 months, no auto-charges. Dedicated with IPMI from $39.20/mo. Crypto invoice, no KYC.

Launch now

One transfer, exact amount: the match the watcher runs

The least intuitive part of a working GRAM payment is how rigid the credit is. There is no minimum-plus buffer, no rounding band, no sweep that harvests whatever happens to arrive. The invoice stores one number and waits for exactly that number, delivered in one transfer, to one address that was generated for your order and nobody else's. Generosity and carelessness look identical to the watcher: both are mismatches.

The address is single-use by design. When the invoice opens, the watcher binds that UQ… account to your order and starts polling; when a transfer lands, it compares the printed amount against the received value and checks the transaction against its own history. A transaction ID that already credited another invoice is refused on sight, which is what stops a replayed or duplicated payment from crediting two orders at once.

This is why partial sends stall instead of partially crediting. Send 90% of the figure and the watcher sees a mismatch: the invoice stays unpaid while your TON sits on a valid address, doing nothing. Send the full amount twice and the second transfer has no unsatisfied invoice left to land on. One transfer of the exact figure is not a style preference — it is the matching rule itself, and the whole troubleshooting chapter of this guide is just cases where one of its three parts slipped.

The typical mistake comes from exchanges. Custodial platforms batch withdrawals, deduct network fees from the requested amount, or route value through internal accounting before the chain ever sees the coins. Any of those turns a clean payment into a near-miss that the match correctly rejects. If you fund the wallet from an exchange, withdraw first and then send from your own wallet, where the fee comes off your balance — never out of the middle of the invoice figure.

Keep the transfer in your own hands and the match runs itself, every time, at a cost of one network fee. The next section covers the place where wallets put that rule at risk in practice: the send screen, where native TON and a jetton sit one tap apart. Three numbers summarize the discipline: 1 transfer, 1 address, 100% of the printed figure.

Tonkeeper works; the send screen is where it fails

Tonkeeper is fine for this payment, and so is any TON wallet whose keys you control. The failure mode is not brand-specific — it hides one tap deeper, on the send screen, where the wallet asks what kind of asset you are moving. Native TON and a jetton sit behind nearly identical screens in several popular wallets, and the invoice credits exactly one of them. Reading that screen carefully is the entire skill.

Before confirming, read the screen the way the watcher will. The asset field should read TON, plain — not a token name, not a ticker with a contract badge beside it. The recipient should be the UQ address printed on the invoice, pasted whole, never retyped from memory. The amount should equal the printed figure to the last decimal. Tonkeeper shows the network fee as its own line; that fee rides on top of the payment and never touches the matched amount, which is exactly how the watcher expects it.

Timing numbers help you stay calm while the chain does its work. TON credit lands in 1–3 min after the transfer confirms, and the box images about a minute after that — the checkout-to-SSH window runs near 60 s once the invoice clears. None of that sequence starts until the transfer is actually on-chain, so a wallet still showing a pending broadcast is not a hosting problem yet. Check the wallet's history view, and then the explorer, before assuming the chain dropped anything.

The classic accident is a swap-first habit. Some users buy a token first and then send the token, because the screen labels it with a TON-adjacent name; others paste the invoice address into a comment field and dispatch to an old contact. Both produce a valid, permanent transfer to the wrong destination. If the send screen offers a memo or comment line, leave it empty — the invoice needs no memo, just the amount at the address, once.

Get the screen right once and the rest is waiting, however the wallet dresses itself. Wallet apps differ in chrome and vocabulary, not in the underlying rule: native asset, printed address, printed amount, single transfer. Every payment story that ends in a support ticket on this rail starts with one of those four fields being almost right, which is another way of saying wrong.

The 2018 Gram is a name collision, not a network

The strangest fact on this page is that the most famous asset ever named Gram cannot be sent anywhere at all. Telegram's Gram was cancelled along with its original project back in the 2018 era; it never launched as planned, so it has no ledger to live on and no balance to move. An invoice cannot accept it for the same reason a cash register cannot stock a currency that was never printed.

What exists instead is The Open Network — TON — an independent chain with its own validators, its own whitepaper, and accounts that start with UQ. The checkout's Gram (TON) entry points at that chain's native asset and at nothing else. The name carries the marketing history; the asset carries the payment. Search engines blend the two constantly, which is why so much GRAM-related content you will find is noise about a token that no longer exists in any payable form.

The collision gets worse one layer down. Token issuers have deployed ERC-20 contracts named GRAM on Ethereum and jettons on TON itself, none of which are the chain asset and none of which connect to this invoice. The reference implementation of the chain lives at github.com/ton-blockchain/ton, and nothing in it mints a token called Gram; the official guide to moving assets, ton.org asset processing, draws the native-versus-jetton line the same way this page does. If a wallet's asset search offers several GRAMs, every single one of them is the wrong answer.

The mistake to avoid is resolving the ambiguity by vibe — sending whichever asset the wallet suggests first because the name matches the invoice label. Names match; assets do not. The resolution is mechanical instead: the invoice prints a UQ address and an exact amount, and only the native chain asset moving between those two points satisfies the watcher. Every other asset is a support ticket with a delay attached, and the ticket is not guaranteed to end well.

One rule absorbs the whole history lesson: if the asset has a contract address, it is a token, and this invoice does not read tokens. The chain asset has no contract — only the chain. Hold onto that sentence and the 2018 saga, the jetton zoo and the Ethereum impersonators all collapse into the same non-event, which is where they belong.

What a fast chain cannot speed up

The counter-intuitive part of TON's speed is how little of the total clock it owns. Credit lands in 1–3 min, and that is genuinely quick — but the VPS it buys images on the same clock as every other coin, roughly a minute of provisioning after the invoice clears, and the payment-to-SSH window lands near 60 s counted from the credit. The chain is the fastest segment of the pipeline by a wide margin, and still the shortest one.

Dedicated metal is where expectations need the hardest reset. Racking takes 2–4 h on the Forge through Monolith tiers — physical installation, cabling, and IPMI credentials handed over last — and no coin shortens it by a single minute. If the schedule is tight, size the urgent half as Core at $4.80/mo annual-eff — the $6 list on the annual term — and let metal take its hours; the invoice engine treats all nineteen coins identically once the credit lands, so GRAM buys no priority queue and no fast lane.

The honest comparison, then: coin choice moves the clock only up to the credit. A two-block Bitcoin confirmation can stretch toward an hour under mempool pressure; Monero's ten blocks typically clear inside twenty minutes; TON lands within one to three. After the watcher fires, every rail collapses into the same provisioning timeline, because provisioning does not know or care which asset paid. Pick a chain for its payment properties — fees, opacity, where your funds already sit — never for a provisioning edge that does not exist.

The mistake is blaming the host for an exchange. The slowest link in a stalled GRAM payment is almost always the sending side — a custodial platform batching withdrawals on its own schedule or holding amounts behind an internal review. The chain itself has no queue for you to wait in once the transfer is broadcast. A ticket containing the transaction ID answers the question in minutes; a resent transfer to the single-use address creates a second problem while the first one is still resolving itself.

Set the clock correctly at order time: minutes for a VPS, hours for metal, and the chain's speed matters only at the front of that sequence. Every ETA you will ever read about this checkout is written in those three units, and keeping them separate is most of what experienced buyers mean when they say a payment went smoothly.

Geography rides the same invoice at the same list price

A coin and a city are independent dropdowns, and GRAM buyers often assume the second one costs extra. It does not: one list price covers all twelve flags — Reykjavik, Zurich, Amsterdam, Frankfurt, Stockholm, Helsinki, Oslo, Luxembourg, Tokyo, Singapore, Hong Kong, Taipei — and the locations table prints each site's tier, uplink and governing law. The coin you pay with changes nothing about which cities are on the menu or what they cost.

The published exceptions are on the location side, not the coin side: Iceland carries a +20% premium — a Tier IV site with 10 Gbps of uplink on FICIX peering and no mandatory data retention law — and Zurich carries +35% for its Tier IV, Swiss-law site on SwissIX. Every other flag sits at base list. Tokyo, Singapore and Hong Kong are the round-trip picks for Asia-Pacific audiences; Taipei adds TWIX to the map; none of them register which of the nineteen coins settled the invoice.

Law is the part you cannot see in a latency map. Iceland's no-mandatory-retention stance, Zurich's judicial overlay on data requests, Singapore's PDPA, Japan's APPI — those read like footnotes until a workload makes them the whole point of the order. If speech process is the reason for the purchase, Reykjavik is the pick; if the audience sits in Asia, the eastern flags win on round-trip time; if GDPR posture matters to your users, the EU sites price the same as everywhere else on the map.

Three mistakes recur here. Buyers multiply the premium onto the whole order when it applies to the list side only, and others skip the premium sites entirely assuming the markup is hidden — it is printed, on the page, before you owe anything. The third: choosing a city for its legal posture while paying with a chain asset that reads in public on an explorer, then filing support requests from a named email. Geography, coin and contact identity are three separate decisions; collapsing them into one is how postures get wasted.

Decide the flag on law and latency, decide the coin on payment properties, and keep the two choices independent the way the invoice does. The dropdowns are adjacent on the same page precisely because neither one constrains the other — a small piece of interface design that encodes most of the jurisdictional logic of this checkout.

Paying for metal with GRAM: same engine, slower clock

Dedicated servers bill on the same invoice engine as VPS slices, and GRAM pays them identically. The lineup runs Forge through Monolith — $49–$149 list, $39.20–$119.20 on the annual term — with IPMI on every tier and DDoS filtering included. The coin changes nothing about the hardware, the disk pairing, the bandwidth or the console hand-over; it is one dropdown on one page, the same as it is for a slice.

What changes is your calendar. After the credit, a VPS images in about a minute; metal takes 2–4 h of racking before the IPMI credentials land. Book Forge — a Ryzen 9 9950X with 16c/32t, 32 GB DDR5 and 2×1 TB NVMe — at $39.20/mo annual-eff against the $49 list when the workload needs local disks and the memory headroom a slice cannot carve out. The dedicated tiers above it scale cores, ECC and port speed; none of them scale faster than the datacenter floor.

The IPMI hand-over deserves a sentence of respect from first-time metal buyers. Every dedicated tier ships with out-of-band console access, which means the OS install, the disk layout and the rescue work all run through a screen that does not depend on the network stack you are about to configure. For a first box, budget an hour with that console after the rack: pairing the two NVMe drives as RAID 1 is the standard first stop, and it is minutes of work once you are actually looking at the screen.

The recurring error is ordering metal for tonight's demo. If the timeline starts in minutes, a VPS tier is the honest answer; if the timeline tolerates half a day, metal wins on disk, memory and kernel control. The second error is picking the coin as if it were a priority flag — GRAM does not jump the racking queue, and neither does anything else. Support ordering runs on ticket timestamps, not on payment rails, which is exactly how it should be.

Pay metal with GRAM happily; just read the ETA on the page as racking hours, not blockchain minutes, and let the IPMI screen be the first door you walk through. The economics — one invoice, annual discount off the USD total, live-rate conversion — are identical to the slice you may already be running, which makes scaling up a payment question you already know how to answer.

Stalled payment: the queue is on the sending side

A GRAM transfer that has not credited within ten minutes almost never means a chain problem. TON confirmations run in 1–3 min; the far more common holder is the sending platform. Custodial exchanges batch withdrawals, hold amounts above internal thresholds for review, or sit on the request until a manual queue drains — and none of that machinery is visible from the invoice page, which is why the wait feels like a hosting failure when it is not.

Diagnose in order. First the wallet: does its history show the transfer broadcast and confirmed? If the wallet itself still shows pending, the hold is there, and only the wallet operator can release it. If the wallet shows confirmed but the invoice still reads unpaid, give it the credit window, then open a ticket containing the txid. The transaction ID lets the watcher search the chain directly — the same getTransactions-style lookup any explorer runs — and answer in one pass whether your payment arrived and what it looked like.

What not to do matters more than what to do. Do not resend blind: the invoice address is single-use and already bound to your order, so a second transfer either duplicates a payment that is about to credit or arrives after the invoice has closed. Do not open five tickets; one message with the txid is the whole protocol. Recovery for a wrong-asset send follows the same path, just slower, because a human has to trace the token rather than query the native ledger.

The systemic mistake is timing: paying inside the last minutes of a deadline from a platform that queues. If the box is needed within the hour, fund a self-custody wallet ahead of time and pay from it, so the exchange's schedule never sits on your critical path. An over-sent amount, once the cause is established, comes back as a ticket case rather than a resend case — the distinction that keeps one incident from becoming two.

Ticket with the txid, wait for the watcher's answer, and keep the resend instinct on a leash — single-use addresses turn impatience into a second case. Every stall on this rail decomposes into a wallet question, a chain question or an exchange question, in that order, and only the third one involves anyone else's schedule at all.

Where GRAM sits among the other eighteen coins

Nineteen coins bill identically on this checkout, which frees the choice to be about the rail rather than the price. GRAM (TON) earns its slot for Telegram-adjacent users: if your funds already live in a TON wallet, the payment is one transfer on a rail you already operate, with no conversion stop in between. There is no discount for it and no surcharge — the invoice converts at the live rate exactly like every other asset on the page.

The neighbors matter for different reasons. Monero is the opaque-by-design option — RingCT conceals amounts and one-time stealth addresses conceal recipients, per the RingCT paper — when linkability is the concern. USDT holds the dollar figure still across the confirmation window on 7 networks. XRP quotes in ledger-native dollars. BTC is the universal one with the permanent public trail. The general flow is walked at buy a VPS with crypto, and the privacy-first case at pay with Monero.

A practical rule: choose by where the funds already sit, then by what the payment should reveal. Money on a TON wallet and a GRAM invoice agree; moving BTC from a cold wallet to pay in TON means an extra conversion, an extra fee and an extra public trace at some swap desk. For a first order, the cheapest path is usually the shortest one from your existing balances to the printed figure — the rail you do not have to visit is the cheapest rail you will ever use.

The mistake is choosing the coin as a statement rather than a rail — picking the most private option while publishing a named project on the box, or the most public one for a personal endpoint. Chain properties are real but bounded: the checkout adds no identity fields, signup being token or email plus a 12-character password, and it never screens wallets or scores coin history. What links your name to the box lives mostly in your own habits and in the exchange you funded from, not in the invoice. For the account-side requirements, the no-KYC guide holds the details.

GRAM is the right answer when TON is where your money already lives. Otherwise the deciding questions run opacity, stability, starting location — in that order, and in that order they are answerable in under a minute, which is about how long the whole payment deserves to take once it is on its way.

The GRAM payment, in the order to run it

Run the whole sequence as one sitting. Pick the plan and the term first — the annual term takes 20% off the USD total before conversion, quarterly takes 10% — then the location, then open the invoice from the deploy page and choose Gram (TON) from the coin list. The page prints the exact figure and a single-use UQ address; from that moment the amount, the asset and the destination are all fixed, and everything you do next either satisfies them or does not.

Send next, from a wallet whose keys you hold: one transfer, native TON, exact amount, network fee on top. Tonkeeper or any TON wallet works as long as the send screen shows the native asset rather than a jetton badge. Credit lands in 1–3 min; a VPS then images in about a minute, which puts the payment-to-SSH window near 60 s from the credit. The box answers before most card payments would have finished a 3-D Secure round-trip, and it asked for nothing but the address along the way.

For metal, extend the same timeline by hours: Forge through Monolith rack in 2–4 h with IPMI credentials arriving last, and the coin neither shortens nor lengthens that. If anything stalls, the order of operations is wallet history first, credit window second, ticket with the transaction ID third — and never a blind resend to the single-use address. The habit of blaming the chain first is the most expensive assumption on this page, because it sends people into resends that create the very mess the ticket would have cleaned up.

Close the loop with the choices that outlast the payment: the location's law, the term's discount, and the workload sized honestly to the tier. The sibling walks share this same skeleton — Monero, USDT, Bitcoin — so the second order you place on this checkout runs faster than the first one did. For the dedicated side specifically, dedicated servers with crypto and Monero on bare metal carry the metal-level detail, from racking to the IPMI console you will meet first.

Is this the Telegram Gram token from 2018?

No. The catalog coin is TON — The Open Network — whose account addresses start with UQ. The 2018 Telegram token was cancelled and cannot be sent anywhere; jettons and ERC-20 contracts named GRAM are different assets entirely, and none of them will credit a VPSbit invoice. Match the native chain asset, and the name collision stops mattering.

Does Tonkeeper work for this invoice?

Yes, provided the send screen shows native TON going to the printed UQ address at the exact amount, in one transfer. A jetton screen in the same wallet is the wrong asset and will not credit; leave the memo field empty and let the network fee ride on top of the amount rather than out of the middle of it.

Is there any KYC step before paying?

No. Signup takes a token or an email plus a password of at least 12 characters — no passport, phone or card anywhere in the flow. TON itself is a public chain, so the transfer is readable on an explorer, but the invoice and the account carry no identity fields to link the payment to a person.

Can I pay for a dedicated server with GRAM?

Yes. The same invoice engine bills Forge through Monolith, $49–$149 list and $39.20–$119.20 on the annual term. Racking adds 2–4 hours after the credit and the IPMI credentials arrive last; the coin changes neither that clock nor the console hand-over, so plan the calendar in hours rather than chain minutes.

What happens if the transfer stalls?

Wait out the sending side first — exchange withdrawal batching is the usual delay, and it runs on the platform's schedule, not the chain's. If the invoice still shows unpaid after the 1–3 minute credit window passes comfortably, ticket the transaction ID rather than resending; addresses are single-use, and a second blind transfer only forks the case in two.

Primary sources

More ways to pay

Pay with Bitcoin · Pay with Monero · Pay with USDT (TRC-20/ERC-20/SPL) · Pay with Litecoin · Pay with Ethereum · Pay with Solana · Pay with x402 · Pay with Dogecoin · Pay with XRP · Pay with TRX · Monero on bare metal · Dedicated servers with crypto · Pick the jurisdiction first.

Ready to launch?

Build the box — VPS or bare metal — create the password, pay the invoice that follows.