Refund Policy

A proxy refund policy defines when a buyer can recover a payment, how much can be returned, and where the money will go. It forms part of the Provider Policies and applies to particular products, payments, and periods of use.

For the buyer, a refund promise reduces the risk of trying a service: pay for proxies, test them on the websites that matter, and leave if the result is unsuitable. But the service starts being consumed between purchase and the refund request. Software downloads pages, addresses are used, and another billing period can begin. These events can affect the decision even when the customer has obtained no useful result.

In the second half of 2026, published policies include guarantees for a first purchase, short evaluation windows, refunds limited to service faults, and separate usage caps. A headline offering “14 days to request a refund” therefore describes only part of what will happen to a particular order.

What a refund guarantee covers

A money-back guarantee can allow an eligible buyer to leave because the service did not meet their needs. Conditions still apply: a deadline, a usage limit, an accepted payment method, and sometimes no previous free trial. Meeting one condition does not replace the others.

Another model limits refunds to technical problems for which the provider is responsible. The supplied proxies might be unreachable, for example, with support unable to restore them or provide a replacement within the stated period. Under this policy, changing your mind is insufficient.

Some purchases are excluded from the standard refund program, while other cases are decided at support’s discretion. A goodwill refund resolves an individual case without promising the same treatment for every subsequent order. A business contract, a paid trial, and an ordinary dashboard purchase can each have different terms.

Consider a customer who buys 100 proxies to collect prices. The addresses open websites, but the required store returns an access denial. The package is useless to the buyer; the provider sees a working network. If the purchase qualifies for a satisfaction guarantee, the request can be considered under that guarantee. If refunds cover network faults only, the dispute turns to what was promised about access to this store.

Refund windows and usage caps

A policy can limit both elapsed time and consumption. One published program, for example, allows 14 days for a first purchase of a service type and caps usage at 20% of the plan or 1 GB, whichever comes first. This is a particular offer with additional exclusions.

That rule produces different testing allowances:

Another published policy considers requests within 48 hours, with usage below 1 GB for proxy server and static residential plans or below 0.1 GB for rotating residential plans, and fewer than 1,000 proxies used. Three measurements matter here: elapsed time, transferred data, and the number of addresses used.

A buyer can meet the deadline and still exceed the usage allowance. If software downloads CAPTCHA pages, access-denied responses, and repeated requests through the proxy, the transferred data can count on a plan billed by bytes. The business task has failed while the usage meter keeps increasing. A product billed by successful results can apply different accounting rules.

The clock also starts from a specified event. If the policy names the payment time, running the software for the first time a day later does not restart it. The request deadline and the period allowed for support to fix a fault can be separate. Waiting for a reply does not automatically extend the refund window.

Which problems qualify

An unreachable proxy, a block on one website, and an application setup error need different explanations. A proxy entered with incorrect credentials is not necessarily faulty. An address rejected by a website on reputation grounds may carry traffic to other destinations successfully. A CAPTCHA does not establish that the network is down.

A working connection does not settle every promise made by the seller, however. If the order specifies a country, an IP type, a protocol, or access to a particular resource, a mismatch concerns the purchased service. Some published policies explicitly tie suitability to the purpose stated at purchase. Testing a completely different website then does not answer the customer’s complaint.

The same distinction matters for speed. A defined throughput commitment differs from a general claim that proxies are “fast.” Poor performance can arise from server load, the network route, the target website’s limits, or the customer’s software. A slow download alone does not identify the cause, but the advertised service specifications still matter.

Some policies provide for repair or replacement first. If ten addresses in a list of 100 fail to meet the agreed conditions, support may offer to replace those ten. Whether the buyer can demand a full monetary refund instead follows from the purchase terms, rather than the number of failed connections alone.

How the proxy product changes the dispute

Network type, access arrangements, and billing can overlap. A residential proxy can rotate; a datacenter proxy can be shared or dedicated. These characteristics help identify what the customer purchased and which problem needs to be assessed.

Residential and mobile proxies

For a data package, consumption is a major factor. For a mobile device or connection rented for a period, the time for which access was supplied also matters. The word mobile alone does not explain how a refund will be calculated.

Even mobile offers can have different caps: published examples include 500 MB for one service model and 100 MB for another. Some consumption also does not automatically rule out a full refund. One published pay-as-you-go (PAYG) program permits a full refund subject to conditions including a request within three days of the first payment and usage of no more than 20%.

Availability of the required exit is another concern. If a service was sold for a particular country and that location cannot be accessed, the issue concerns delivery of the order. If exits are available but one website rejects them, the terms concerning target access become relevant. An unsuccessful task and the remaining data balance are separate parts of the case.

Shared and semi-private proxies

With shared access, several customers use the same exit IP. An offer sold as semi-private may limit the number of tenants, for example to three or five. That number describes the particular offer.

A website can receive those customers’ requests from one address. Another tenant’s activity can affect later access even though the buyer has changed nothing. A refund dispute therefore depends on what the seller promised: network availability, defined sharing restrictions, or suitability for a stated resource.

Support may offer an IP replacement in this situation. Shared access itself establishes neither a ban on monetary refunds nor a mandatory “one free replacement” rule. Poor performance on a shared plan also does not, by itself, establish that the provider has oversold capacity.

Dedicated addresses

For a dedicated package, exclusivity and the condition of the addresses at delivery matter. A buyer might receive an IP already blocked on the required website, or encounter a ban after starting to use it. These situations raise different questions about the order’s quality.

If the terms exclude replacement of addresses blocked because of the customer’s actions, intensive use can affect the outcome. The time at which a ban appeared does not prove its cause, however. The assessment needs to account for the sales terms, earlier availability, and the circumstances of the block.

A listing in an email spam database, a social network’s rejection, and an unreachable proxy are different problems. An address can satisfy one description without satisfying the others. Calling an IP “clean” without naming the check leaves room for very different expectations. The meaning of private also comes from the offer: the label does not always promise a single user.

Rotating proxies

The address-changing function itself can be the subject of a complaint. An offer might promise a new exit for each request, yet the exit remains unchanged when the corresponding connection method is used. This differs from a customer using session settings that deliberately keep one address.

Repeated addresses are another issue. A selected city or network operator may have a small pool, so working rotation can return an IP that has already been used. An address change does not promise that the address will never repeat. If the seller separately guarantees a number of unique IPs or the size of an available pool, that commitment becomes part of the service assessment.

Unused service and partial refunds

The remaining allowance in a dashboard shows what is still available to use. Whether it can be converted back into money follows from the purchase terms.

If a customer rents a fixed IP list for a month and stops after a week, the addresses may have remained allocated throughout that time. The plan supplies access whether or not the customer’s software is running. A data package is different: an unused volume remains visible, but it can also have an expiry date and refund restrictions.

A partial refund, where available, can account for remaining time, consumption, resources already supplied, and the terms of a discount. The amount therefore need not equal a simple fraction of the original payment. Promotional credits and bonus data are not money contributed by the buyer either.

Adding funds to an account requires separate consideration. The money might not yet have purchased a particular service, but that does not guarantee unrestricted withdrawals. A policy can treat deposited funds, compensation credits, and bonuses differently. When changing plans, a remaining allowance may be applied to another purchase within the same service.

First purchases and later charges

A guarantee can apply once per account or once per product type. These are different conditions. A free trial before purchase, a paid test package, and a negotiated business contract may also fall outside the ordinary refund program.

Canceling a subscription, disabling renewal, and refunding an existing charge are separate operations. Turning off the next payment does not reverse the previous one. Immediate cancellation, in turn, can end current access or trigger an account-credit calculation.

Renewal rules do not necessarily repeat first-purchase rules. Some offers exclude later payments; others provide a separate window after an automatic charge. Published mobile terms, for example, distinguish 72 hours for certain automatic renewals from 24 hours for manual renewals, with additional usage conditions.

An automatic balance top-up is another case: the charge adds funds without necessarily purchasing another month. Duplicate charges and incorrect amounts concern payment accuracy. Their meaning differs from a complaint about proxy quality even when the same support team handles them.

When changing plans, two relevant questions are whether the remaining allowance transfers and whether the original guarantee continues. Seeing a different package in the dashboard does not, by itself, create a new first purchase or refund window.

Money, account credit, or replacement

A message saying that the purchase has been compensated can describe several outcomes:

The distinction matters to someone who wants to leave. Account credit supports continued use of the provider but may not be withdrawable. Replacement keeps the purchase in place with different addresses. Extra days have value only if the repaired service suits the task.

Compensation for downtime can also be calculated separately from a purchase refund. A service-level agreement may award credit for failing to meet a stated availability commitment. That award should not automatically be equated with the cost of the whole package.

Verification, restrictions, and misuse

Documents requested after payment

A difficult situation arises when payment has been taken but access still depends on identity or business verification, known as KYC. The customer might decline to submit documents, fail verification, or be refused service by the provider.

These cases have different causes. The payment outcome depends on activation terms, what has already been supplied, and the treatment of unused funds. A successful payment does not mean verification is complete, and a document request alone does not explain whether a refund will follow.

Service territories and sanctions

Access can be restricted by a provider’s own rules, a payment intermediary’s requirements, or applicable sanctions. Mentioning Russia, Belarus, Iran, or a particular territory does not fully describe which customers and transactions a restriction covers. Residence, current location, parties to the payment, and other circumstances can matter.

For the buyer, refusing service, holding a transaction for review, and legally freezing funds are especially important distinctions. Under the rules administered by OFAC, the US sanctions authority, blocked property remains owned by its owner, but transfers or refunds require the relevant authorization. In other cases covered by that system, a prohibited transaction is rejected and funds are returned to the sender.

A closed account and a frozen payment therefore describe different events. One case can be resolved under the service’s terms; another can involve legal restrictions binding a party to the transaction.

Abuse and termination

Service terms can provide for termination without a refund following violations such as spam, password attacks, DDoS, and other prohibited activity. Being within the initial guarantee period does not remove these exclusions.

A complaint, a temporary suspension for investigation, and a decision that a violation occurred are separate stages. Receiving an abuse report does not, on its own, establish what happened or which customer’s activity was involved. That distinction is particularly relevant when an IP is shared by several tenants.

Payment method and receiving the money

The payment method can change the refund conditions for the same product. Cards, bank transfers, payment intermediaries, and cryptocurrency follow different processes.

Some services expressly exclude refunds for BTC, USDT, LTC, and other cryptocurrency payments. Others publish terms allowing repayment in the same cryptocurrency, with a stated exchange-rate rule and fees deducted. An irreversible original transaction does not technically prevent a new transfer; the question is whether the seller offers it and whether it is permitted in the particular case.

A promise to refund the full purchase price can also contain an exception for payment costs. If the terms deduct receiving, processing, or return-transfer fees, the amount received will be lower than the original payment. Exchange-rate changes can also affect what appears in the buyer’s own currency.

Review, approval, sending the refund, and crediting the recipient are separate stages. A support message confirming approval does not mean the bank has already displayed the money. Card refunds normally follow the original payment route, with handling dependent on the payment network and bank. In some cases, the original charge disappears from the statement instead of a separate credit appearing.

A bank dispute, or chargeback, follows a separate process. The bank and payment network consider the dispute’s grounds and the information submitted; a support refusal guarantees neither success nor failure. The seller’s commercial policy is also not a complete account of the buyer’s mandatory rights in every jurisdiction.