FineProxy FAQ

FineProxy is a proxy server rental service operating on its own infrastructure. We use modern AMD-based server hardware deployed across dozens of data centers worldwide. Our network is built with high-capacity uplinks of up to 80 Gbps, ensuring stable performance even under high load. All proxies are provided from our own IP pools (we operate as a direct provider) and remain active for the entire paid rental period.

We offer datacenter proxies in packages, dedicated proxies assigned to one customer, static ISP proxies, and rotating proxy products.

Supported protocols are HTTP, HTTPS, SOCKS4, and SOCKS5. UDP can be added to any plan except rotating proxies, either while ordering or later from the active service page. When added later, the charge is prorated for the remaining service days.

In most cases, IP addresses won’t overlap, because each order is provisioned separately.

However, with package proxies, a small amount of overlap is theoretically possible, since IPs may be shared among a limited number of customers. The probability of overlap can increase with larger package sizes, but in practice it is uncommon.

If you need guaranteed IP uniqueness, dedicated proxies are the right option.

A package datacenter order can include one or multiple countries as long as the total order contains at least 100 IP addresses. Dedicated proxies are excluded from this package minimum.

Country Mix plans let you exclude unwanted countries and adjust each country’s IP address quantity before or after purchase, subject to the plan and available inventory.

Dedicated proxies are IP addresses assigned exclusively to one customer for the entire rental period. These proxies are not used by other customers during the lease term and remain unchanged throughout the paid period unless you request a replacement or renewal. Dedicated proxies are a good fit when you need a strictly exclusive, static IP.

Package proxies are delivered in bundles with a fixed number of IP addresses. Depending on the plan, individual IPs in a package may be used by a limited number of customers (typically up to 3–5) at the same time. This model is designed for scenarios where IP volume matters more than per-IP exclusivity.

All package proxies are still our own IP addresses running on our infrastructure. These are not free proxies and not sourced from scraped third-party lists — they are controlled IPs available only within our proxy service.

The key difference is the exclusivity model.

With dedicated proxies, the IPs are assigned to a single customer and are not shared during the rental period. The IPs remain tied to your order.

With package proxies, IPs are provided as a bundle and may be shared by a limited number of customers (typically up to 3–5). This allows us to offer larger IP volumes at a lower price, but it does not guarantee exclusive use of every individual IP. This model works well when scale is more important than having strictly dedicated IPs.

ISP proxies are static residential IP addresses that belong to ordinary home internet service providers but are hosted on servers in a data center.

This gives you the strengths of both proxy types. From the data center you get high speed, a static IP address that never changes, and stable uptime. From residential proxies you get trust: websites see the ASN of a regular internet provider rather than a hosting company, so these IPs are far less likely to be flagged or blocked.

You can find plans and prices on the ISP proxies page.

API credits are the internal billing units used to measure how much load your requests generate. Every plan comes with a fixed number of credits, and every outgoing request deducts a certain amount based on how complex the target webpage is.

A simple request that loads only HTML or JSON typically costs 1 credit. Requests that trigger multiple parallel assets — such as scripts, images, CSS files, fonts, APIs, and tracking requests — consume more credits. Each asset counts as a separate downstream request.

These numbers are approximate and vary per website:

• 1 credit — simple API-style request, no browser, no JS
• 2–3 credits — no-JS browser request (basic DOM, no rendering)
• ≈10 credits — full headless browser with JS rendering, typical DC proxy load
• 10–15+ credits — heavy pages with many scripts, trackers, ads, analytics, or large asset bundles

Different websites have different DOM structures, JS bundles, and asset chains. One site may load only a few files; another may trigger 30+ downstream requests per page.

How to calculate your real consumption? You can easily estimate credit usage before committing to a plan:

• Open the page you want to scrape.
• Check how many parallel requests it makes using a tool like RequestMap
• The number of loaded assets ≈ number of credits per page load.

This gives you an accurate prediction of how many API credits your workflow will consume. Any request that reaches our infrastructure consumes credits — including responses like 404, 429, 503, or timeouts. Even though the target site returned an error, we still processed and routed the request.

No. Location selection is available only at the country level. City-level targeting within a country is not supported. IP addresses are assigned automatically from the least-loaded networks within our infrastructure.

You can top up your dashboard balance by credit card, Apple Pay, or cryptocurrency — BTC, ETH, LTC, USDT TRC20, and USDT ERC20. Card and Apple Pay payments are processed by Paddle and Stripe.

All top-ups go through payment processors only; we do not accept direct cryptocurrency transfers. Available options depend on your region.

Proxies are delivered automatically immediately after successful payment. With standard plans, access to proxies appears in your personal account without waiting and without any manual processing. We do not send proxy details by email.

The only exception is custom or non-standard orders placed through the ticket system. In such cases, proxy delivery may take up to 24 hours.

The standard rental period is at least 30 days. A paid 48-hour proxy term is also available for every product except rotating proxies.

Choose any non-rotating proxy product, configure the countries, IP quantity, and add-ons, then select 48 hours. The catalog calculates and shows the full price for the selected configuration before payment.

If you are happy with the results, you extend the same service to 30 days: the login, password, and IP address list stay the same, so nothing needs to be reconfigured.

Please note: 48-hour orders are not eligible for a refund.

Your dashboard has a single internal wallet. Top it up manually or enable automatic top-ups — every service and add-on is paid from this balance, with no separate invoices to fund for each purchase.

All dashboard top-ups are processed only through payment processors. We do not accept direct cryptocurrency transfers.

Ordering a service for 3 months upfront gives you 10% off, 6 months — 15% off, and 12 months — 20% off. The discount is applied automatically when you select the term at checkout.

All proxies remain active for the entire paid rental period and do not require mandatory replacement. IP addresses stay active as long as you continue to renew your service.

Yes. A scheduled IP list replacement is free: you can run it yourself in the dashboard once every 8 days.

If you do not want to wait for the scheduled replacement, an unscheduled replacement — full or partial — is available. This is a paid option: $10 per operation, deducted from your dashboard balance.

Rotating proxies do not need replacement: their addresses change automatically.

Open the active service → More Actions (top right) → Refresh IPs. Here you can refresh the entire list, replace only selected addresses, or replace addresses from a specific country.

Where your plan and available inventory permit, you can also increase or decrease a country’s share in your package while refreshing. A scheduled replacement once every 8 days is free; an unscheduled one costs $10 — see “Can I replace IP addresses before the rental period ends?” for details.

All addresses in your service are listed on the active service page under IP Addresses. You see the full list with each address’s country, and the country filter helps you quickly pick out the ones you need.

You can export the list in two ways:

• Plain TXT file — select the countries you need and download only their addresses, one proxy per line.
• Custom builder — define your own line format with the {ip}, {port}, {login}, {password}, and {scheme} variables so the file fits your software without manual editing.

For ports and builder details, see “Which ports and proxy-list formats are available?”.

Use the port matching your protocol and authentication method:

Builder variables:

• {ip} — proxy IP address
• {port} — connection port
• {login} — service login
• {password} — service password
• {scheme} — proxy protocol

Arrange them with any separators or text for an arbitrary output format.

These are service credentials, not your dashboard login. Change the login on Connection and the password under Management → Password.

This helps when migrating naming schemes or updating many existing configuration files; update those files with the new values before switching workloads.

Go to active service → Management → Additional IPs. Add IP addresses for one or multiple countries without canceling.

The price is prorated for the remaining days.

Go to active service → Management and select “Cancel this service” at the bottom. Unused value returns to your wallet.

If canceled within 24 hours, you may request a refund to the original method through tickets, subject to the refund policy.

We do not impose artificial speed limits at the plan level. Actual performance depends primarily on external factors, including client-side limitations, software configuration, network quality, and the behavior of target websites.

In real-world conditions, speeds on remote servers can reach up to 700 Mbps. On typical home network setups, speeds are usually around 200–300 Mbps. However, we do not guarantee a minimum speed, as it cannot be consistent across all networks and usage scenarios.

We do not use outdated proxy formats limited to 10–20 Mbps. Speed parameters are specified per individual IP address, not averaged across an entire package.

Data transfer is unlimited. For bandwidth, we recommend staying within the following guidelines, which depend on package size:

• Packages with up to 300 IP addresses — up to 100 Mbps
• Packages with up to 3,000 IP addresses — up to 500 Mbps
• Packages with 3,001 to 5,000 IP addresses — up to 1 Gbps
• Packages with 5,001 to 15,000 IP addresses — up to 1.5 Gbps
• Packages with 15,001 to 25,000 IP addresses — up to 2.5 Gbps
• Packages with 25,001+ IP addresses — up to 5 Gbps

These are not automatic limits — nothing throttles your speed technically. We ask that you stay within these values: if you consistently exceed them, support may open a ticket asking you to reduce the load.

The limit for simultaneous active proxy connections (the technical term is “concurrent connections”) depends on package size:

• Packages up to 999 IP addresses — 1,500 concurrent connections
• Packages from 1,000 to 4,999 IP addresses — 3× the number of IP addresses
• Packages from 5,000 to 14,999 IP addresses — 2× the number of IP addresses
• Packages with 15,000+ IP addresses — the same number of connections as IP addresses

Close connections when each task finishes. Otherwise, a connection can remain active until the 120-second timeout, reducing effective request capacity.

If your workload does not match the statistics, ask your software developer to verify how connections are handled.

Concurrent connections describe how many proxy connections can work at the same time.

Each time your browser, application, or script sends a request through a proxy, a connection is opened. While this request is being processed, the connection remains active and is counted as a concurrent connection. Once the request is completed and the connection is closed, it no longer counts.

It is important to understand that opening a website does not always mean a single connection.

When a page is loaded, the browser first requests the main HTML file. After that, it may load many additional resources such as images, scripts, stylesheets, fonts, and API requests. Each of these requests can open its own connection while it is being processed.

As a result, one website visit may create multiple concurrent connections at the same time. The actual number depends on the website’s structure, the number of assets it loads, and how these requests are handled by the browser or application.

The service automatically enters Slow Mode — a concurrent-connection restriction mode. While it is active, no more than one connection per IP address is allowed.

Its duration depends on violations within 24 hours:

• Up to 2 violations — 5 minutes
• 3 to 6 violations — 15 minutes
• 7 to 10 violations — 30 minutes
• More than 10 violations — 60 minutes

When the restriction period ends, the service automatically returns to normal operation.

Threads are parallel software tasks; connections are active network sessions those tasks open through proxies. One thread can use several concurrent connections.

Typical examples:

• API request — typically 1 connection
• Simple HTTP request (HTML or JSON) — usually 1 connection
• Browser request without JavaScript — 2–3 connections
• Headless browser with JavaScript rendering — about 10 connections on average
• Heavy page with ads, analytics, tracking, and many assets — 10–15+ connections

Each loaded script, stylesheet, image, font, API call, or tracker can create a separate downstream connection.

To understand how many connections you will actually need, measure the load before choosing a plan:

• Open the website you plan to work with in Chrome or Firefox.
• Press F12 (or right-click → Inspect) to open the developer tools and switch to the Network tab. It shows every request the page sends while loading.
• Reload the page and check how many requests run at the same time. The number of resources loading in parallel — images, scripts, stylesheets, fonts, API calls — is roughly the number of connections per page load. The online tool Request Map gives a helpful visual overview as well.
• Multiply that number by the number of parallel threads or tasks in your software, and add a 20–30% margin for peak load.

Websites are built differently: one loads a handful of files, another sends 30 or more requests per page load. That is why you should measure your actual target site instead of relying on averages.

Open the active service page → Statistics. On the chart, switch to the Connections tab: it shows your limit, current usage, and a Slow Mode marker if it was applied.

Yes. If your package’s connection limit is not enough, you can buy additional concurrent connections — either when placing the order or at any time while the service is active.

Everything is done in the dashboard on the active service page: choose the amount you need, and once it is paid from your balance the limit increases automatically, with no support ticket required.

The price is $30 per 1,000 connections for 30 days. If you add connections to a service that is already running, the amount is prorated for the remaining days — you pay only for the rest of the period.

The concurrent-connection limit applies to the service as a whole, not to each allowlisted IP address. If several addresses are allowlisted — for example, your office, a home PC, and a server — they share the common limit by default.

To keep one binding from consuming the connections of the others, you can specify how many connections to allocate to each binding when adding or editing it under Management → IP Allowlist. For example, with a limit of 1,500: 1,000 for the server, 300 for the office, and 200 for the home PC.

The total of all quotas cannot exceed the service limit. You can change the quotas later by editing the binding.

IP allowlisting controls which source IP addresses may access your proxies. Other source addresses are denied.

Access from a newly added source IP address normally activates in about 10 seconds.

Yes. Manage them on the active service page under Management → IP Allowlist.

Up to 5 bindings are free. Each extra binding costs $1 per month and is charged from your dashboard wallet.

No. Using proxies without IP whitelisting is not supported. To access the proxies, your IP address must be added to the allowed list.

If your ISP changes your IP address within the same subnet, the proxies continue to work without constant allowlist updates.

Add your current IP address in the service settings and use login and password authentication. The system automatically determines the allowed subnet range.

For credential-based connections, use:

• HTTP — port 8080
• SOCKS5 — port 1080

For example, with 85.249.3.145, the system may allow 85.249.0.0–85.249.7.255. Access remains available if your address changes within that range. You can check the range with any CIDR calculator, but you do not need to enter it manually.

The API lets software use supported FineProxy account functions. MCP gives compatible AI tools a standard way to use them.

Create keys at Settings → API Keys. The secret is shown once; copy and store it securely. Use separate least-privilege keys for integrations.

Grant only required permissions:

• customer:read — read customer information.
• customer:write — update customer information.
• wallet:read — view wallet data.
• wallet:deposit — manage wallet deposits.
• services:read — view services.
• services:write — create or change services.
• tickets:read — view tickets.
• tickets:write — create or update tickets.
• catalog:read — view products and options.
• usage:read — view usage data.
• wallet:spend — lets automation spend wallet funds.
• webhooks — manage webhooks.

wallet:spend is financial permission: automation with it can spend wallet funds.

A webhook is an automatic notification that FineProxy sends to an address you specify (an HTTPS URL) whenever a billing event happens in your account — for example, wallet activity or changes in service payments. You do not need to log in to the dashboard or keep polling the API: the system tells your service about the event on its own.

What this is useful for in practice:

• Accounting — automatically record payments and top-ups in your CRM or bookkeeping system.
• Service monitoring — get a signal when a service is paid or renewed and immediately trigger your own processes, such as updating the proxy list in your software.
• Alerts — forward notifications to Telegram, Slack, or email so your team learns about events without opening the dashboard.
• Balance control — tie wallet events to your own checks and top up in time so services are never interrupted.

Setup: Settings → Webhooks → enter the HTTPS address where your service accepts requests. A secret is shown once when the webhook is created — save it: your service uses it to verify that a notification really came from FineProxy and not from a third party.

If you do not run your own server, you can receive webhooks through no-code tools such as Zapier, Make, or n8n.

This is caused by IP Whitelisting. The IP address used by the online checker is different from the IP address whitelisted in your personal account, so the checker cannot access the proxies. Our proxies will work only from the whitelisted IP address.

Test in the same environment and software you plan to use:

• Configure proxies in Firefox with the FoxyProxy extension.
• Use SOCKS5 with authentication.
• Test several websites.
• If they still fail, verify the source IP address under IP Allowlist.

Test with minimal load and verify IP Allowlist, especially if your source IP address or subnet changed.

Try other websites. If only one fails, check the blocked websites and services policy. Resources covered by it cannot be unblocked.

If the resource is not blocked, send support the error code, logs, and a screenshot of your proxy settings.

All requests related to technical issues, proxy checks, and non-standard situations are handled exclusively through the ticket system. This allows us to properly track each case and provide an accurate response. The average response time is up to 1–2 hours, while the official response SLA is up to 24 hours.

Our proxies operate with infrastructure-level restrictions that apply to all users and all plans. If a resource falls under these restrictions, access via our proxies is not possible, and unblocking is not available.

• Email services and mail protocols (SMTP, IMAP, POP3) are fully blocked, regardless of use case.
• Government resources, including all .gov domains, are blocked for all countries and plans.
• Google Search results are blocked (other Google services remain available).

The full list of blocked websites and services is available here

If a resource is listed there, it will not work through our proxies.

We guarantee proxy availability throughout the paid period when service terms and limits are followed.

We do not guarantee access to every website or compatibility with a particular service. Results can depend on changing target rules, anti-bot systems, browser profiles, headers, headless settings, and other software configuration.

If particular addresses do not suit you, they can be replaced — the options are described in “Can I replace IP addresses before the rental period ends?”.

Yes. We provide a hassle-free refund within 24 hours after the service purchase. Refunds are processed using the same payment method that was used for the purchase. If the payment was made in cryptocurrency, you will need to provide a wallet address and the network name. All refund requests must be submitted exclusively through the ticket system in your personal account. Before requesting a refund from your account balance, you must manually cancel the active service in the personal account. This is required to comply with procedural and regulatory requirements.

A refund may be declined when:

• The service was used in violation of service rules.
• More than 24 hours have passed since payment.
• The service was purchased for 48 hours; 48-hour orders are not eligible for a refund.

After 24 hours, you may still cancel an active service. Its unused value is credited to your dashboard wallet for future purchases.