Loading page content.
About FleetProxy
FleetProxy exists because buying proxies at scale is a market where the specifications are vague, the meter is invisible, and the sourcing question gets answered with two words on a landing page. We built the network we wanted to buy: metered per byte in public, priced without a negotiation, and sourced from people who agreed to it in language they could read.
What we do
FleetProxy sells access to rotating residential, static residential, mobile, and datacenter address pools, billed per gigabyte or per leased IP depending on the product. That is the whole business. There is no data-as-a-service upsell, no per-seat licence, and no minimum commitment.
Eight proxy products on one account and one credential format. Bandwidth-billed pools are metered per byte in both directions and the meter is visible per hour and per country in the dashboard, because a bandwidth number you cannot break down is a number you cannot dispute.
Purchased bandwidth stays on the account for as long as the account is open. Most of the industry resets the counter at the end of the billing period, which turns a bursty workload into thirty to forty percent waste. That is a policy commitment, not a launch promotion.
Regional weak spots, the success rates we measure, what the acceptable use policy prohibits, and the comparisons where a competitor wins. A vendor that will only describe its strengths is asking you to discover the rest in production.
The commercial terms are the same for everyone and are published in full on the pricing page: volume tiers apply automatically at checkout, there is no code to enter, and there is no published tier that a sales conversation improves on.
Sourcing
Every residential proxy network is somebody's home connection. The only question that matters is whether that person knew, and whether they can leave. Here is our answer, in enough detail that you can hold us to it.
Residential and mobile exits come from peers who opted in through a partner application that states, on the install screen and not in an end-user licence agreement footnote, that their connection will carry other people's web traffic while the device is idle and charging. Consent that has to be discovered is not consent.
Every peer receives something of value for participating — a paid subscription, an ad-free tier, or a direct revenue share. Supply that is acquired for nothing is supply acquired from someone who did not understand the trade, and it is the single strongest predictor of a pool that will be pulled out from under a customer.
A peer can leave from inside the partner application at any time, and anyone can ask us directly whether an address is in the pool and have it removed. Requests go to the support address and are actioned within one business day; we do not require the requester to prove they are a customer.
We do not take supply from compromised hosts, from libraries bundled into applications without a disclosure screen, or from devices whose owner is a different party to the person who clicked accept. Where a supplier cannot show us the consent flow, we do not buy the range.
The gateway records byte counts, request counts, success and failure counts, latency and exit country. It does not record destination hostnames, URLs, headers, or content, and the usage tables have no column capable of holding them. That protects the peer and the customer with the same design decision.
Supplier consent flows are re-checked quarterly and whenever a partner ships a material change to its onboarding. A range that fails the check is withdrawn from rotation before the review is finished rather than after it.
If you believe your connection is carrying FleetProxy traffic and you do not want it to, write to [email protected] with the IP address and the approximate time you observed it. We confirm whether the address is in the pool, remove it within one business day, and tell you which partner application enrolled it so you can uninstall it at the source. You do not need an account with us, and we do not ask for one.
We cannot tell a peer which customer used their connection, because the gateway does not record it — the same design that keeps destination data out of our systems keeps that mapping out too. We hold no third-party sourcing audit and no SOC 2 or ISO 27001 certification. If your procurement process requires one, that is a genuine blocker rather than a preference, and we would rather say so on this page than in month three of an evaluation.
Boundaries
Automated collection of public data is permitted and is what most of our customers do. These categories are not, and an account is terminated for them rather than warned.
The binding text is the Acceptable Use Policy, which also sets out how abuse reports are handled and what we preserve when we receive valid legal process. Report abuse to [email protected].
The network
These counts come out of the location inventory at request time. Coverage is not uniform, and the per-country pages show the pool size for each market so you can check a specific country before you order rather than after.
Country, city and ASN targeting are segments of the proxy username, so one credential serves a scraping fleet, a rank tracker and a research sweep at once. Where a segment names a pool that does not exist, the gateway returns 407 with an x-fleet-reason header instead of silently falling back to another country — a silent fallback is how a German dataset quietly turns into a US one for two days.
We target 99.9% monthly availability on the authentication gateway. That is measured on the gateway accepting and authenticating connections, not on whether a third-party site let a request through, because the second number is not ours to promise.
Europe and North America are our strongest regions on both latency and success rate. South American residential pools are thinner, latency there routinely runs past 800ms, and customers with heavy Brazilian volume often run us alongside a second vendor. We would rather write that here than have you discover it in week two.
We are also not an enterprise procurement vendor. There is no certification portfolio, no contractual SLA with service credits, and no named account manager below volume tiers.
Check pool size by countrySupport
Support is the product for most of the first month. It is staffed by engineers who can read a gateway log, and it is measured on first reply rather than on tickets closed.
Every contact becomes a ticket with a reference of its own, visible in the dashboard and answerable by replying to the email. Threads survive an agent leaving, they are auditable months later, and they do not depend on a third-party script we would have to admit to our content security policy.
Engineers cover the queue continuously rather than during a single timezone's business hours. The median is a measured figure from our own ticket table, not a target, and it is the median first human reply rather than an automated acknowledgement.
A gateway question gets the relevant log excerpt and the x-fleet-reason value, not a template about clearing your cache. If the fault is on our side we say so and credit the bandwidth; if it is in your client we show you the request that proves it.
Volume customers get a named engineer and a shared channel. Everyone else gets the same queue and the same people — there is no support tier sold separately, and no priority lane that quietly deprioritises the accounts that did not buy it.
Reach us
Pick the one that matches what you need. Anything sent to the wrong address is routed rather than bounced, and the contact form opens a numbered ticket either way.
Support
[email protected]Gateway errors, targeting, billing questions, and refunds under the published policy.
Sales
[email protected]Volume above the published tiers, invoicing terms, and trials larger than the standard 50MB.
Partnerships
[email protected]Reseller agreements, supply partnerships, and integrations that need a written agreement.
Community channels, expected response times and the contact form are all on the contact page. The documents that govern the relationship are the Terms of Service, Privacy Policy and Refund Policy.
The free trial is 50MB of rotating residential bandwidth, granted on account creation with no card and no call. It is enough to run a few hundred requests against the sites you actually care about and measure the success rate yourself across 56 countries of live inventory.
No card required for the trial. One trial per account.
