How Many Accounts Per Proxy? A Practical Guide by Use Case

How many accounts per proxy — cover image

Last updated: August 20, 2026

For any account you log into, plan on one dedicated IP per account. Higher ratios are only acceptable for stateless traffic that never logs in, where you size by concurrent sessions instead of accounts. At Thunderproxy we map this to static ISP addresses for persistent identities and rotating residential for stateless volume. What changes the answer is how sensitive the target is to address reuse, how long each session runs, and whether you hold a dedicated IP or draw from a shared pool.

The short answer, by use case

The accounts-per-proxy ratio is not one number. It depends entirely on whether the work involves a login. The table below gives the practical ratio for the most common jobs. For every login-based use case the answer is one account per IP. For stateless work there is no account ratio at all, only a concurrency limit.

Use case Accounts or sessions per IP Proxy type Session mode Why
Long-lived social profiles 1 Static ISP Persistent Fixed identity needs fixed address
Marketplace seller accounts 1 Static ISP Persistent One storefront, one address
Ad accounts 1 Static ISP Persistent High sensitivity to reuse
Crypto wallet dashboards 1 Static ISP Persistent Login-bound identity
Email mailboxes 1 Static ISP Persistent Login-bound identity
Agency-managed client accounts (team-shared) 1 Static ISP Persistent One address per client identity
Checkout or drop tasks 1 per task session Rotating residential Sticky Short-lived, per-task session
Scraping and monitoring, no login Per concurrent session Rotating residential Rotating No session state to protect
SERP and rank tracking Per concurrent session Rotating residential Rotating Stateless queries
QA and geo-testing Per concurrent session Rotating residential Sticky or rotating No persistent identity

Compliance note: platform terms decide how many accounts you are allowed to hold. Proxies only decide how cleanly the accounts you are entitled to hold stay separated on the network. This page is about capacity planning and session hygiene for legitimate account management, not about evading enforcement or terms of service.

Why the ratio matters at all

A shared address is a shared signal. When two accounts you keep separate log in from the same IP, you have told the network they are related, whatever your intent. That link gets stronger when the same device fingerprint sits behind the same address, because the two signals reinforce each other.

There are two more mechanical reasons the ratio matters. First, rate limits are enforced per address, so stacking accounts behind one IP means they compete for the same request budget and trip limits sooner. Second, a rotation mid-flow invalidates the session: if the address changes while you are logged in, the far side sees the session move and can break it. None of this is about hiding. It is about not sending contradictory network signals for identities that are supposed to be independent.

The three variables that set your ratio

Your ratio is decided by three things, in order.

1) Dedicated IP versus shared pool. A static ISP address is one address you hold for the length of the lease. A rotating residential proxy is a pool you address by session id, and even a sticky session is best-effort and bound by a TTL. If an identity must persist, it needs an address that persists, which means a dedicated IP. If the work is stateless, a pool is not just acceptable, it is the correct and cheaper tool.

2) Session length versus rotation. A long-lived login wants the address to stay put for hours or days. A stateless request wants a fresh address as often as possible. Match the session mode to the work: persistent for logins, rotating for volume, sticky only for short bounded tasks.

3) How sensitive the target is to address reuse. Some platforms treat a repeated address as routine. Others treat it as a strong link between accounts. When a target is sensitive, keep the ratio at one and do not share.

Static ISP versus rotating residential: which one per account

Use a static ISP proxy when a persistent identity needs a fixed address. A social profile, a seller storefront, an ad account or a mailbox is the same identity every time it logs in, so it should present the same address every time. A Thunderproxy static ISP address is a 30-day lease with unlimited extensions and unlimited bandwidth, which means the identity keeps one stable address for as long as you keep renewing it, and heavy in-session data does not change the bill. Coverage is limited to a smaller set of countries, so confirm your target region is available. See the static ISP plans for current country coverage.

Use rotating residential when the work is stateless and per-GB rotation is the cheaper and correct choice. Scraping, monitoring, SERP and rank tracking, and geo-testing do not hold a login, so they gain nothing from a fixed address and everything from a large, fresh pool. Rotating residential draws from 30M+ IPs across 155+ countries with country, state and city targeting, and you are billed for the data you move, not for addresses you park. For the difference between rotating and sticky behavior, see sticky sessions.

How to size and price your fleet

Size the fleet to identities for logins and to concurrency for stateless work, then price each with the matching billing model.

Example 1: an agency with persistent client logins (per IP). You manage 10 client accounts that each need a stable identity: social profiles and ad accounts. The ratio is one dedicated IP per account, so you need 10 static ISP addresses.

  • 10 accounts x 1 IP each = 10 static ISP IPs
  • 10 IPs x $1.79 per IP = $17.90 for the 30-day lease
  • Bandwidth is unlimited, so in-session data does not add to that figure

The bill is flat and predictable because it is tied to identities, not traffic.

Example 2: a stateless monitoring job (per GB). You run scraping and rank tracking with no login. There are no accounts to count, only data to move, so rotating residential billed per GB is the correct model.

  • Assume the job moves 200 GB in a month
  • 200 GB x $1.72 per GB = $344
  • You size addresses by concurrent sessions, using session ids in the 1 to 20000 range, not by a per-account ratio

The comparison is the point. For persistent logins, per-IP pricing turns your fleet cost into a simple multiple of identities. For stateless volume, per-GB pricing tracks the work itself, and buying dedicated IPs you never keep logged in would be pure waste.

Common mistakes

  • Sharing one address across accounts on the same platform. This links the identities on the network and makes them compete for the same rate limit. Give each login its own address.
  • Rotating on a login flow. A new address mid-session invalidates the session. Use a persistent or sticky mode for anything you log into, and reserve rotation for stateless requests.
  • Geo mismatch between address and account. A regional account presenting from the wrong country contradicts its own profile. Match the address region to the account with country, state or city targeting.
  • Reusing one address across different platforms for the same operator. Distinct platforms that keep seeing the same address can correlate the operator behind them. Separate by identity, not just by platform.
  • Buying more addresses than concurrent sessions. For stateless work, idle dedicated IPs cost money and do nothing. Size to peak concurrency and let rotation cover the rest.

For the wider workflow around keeping identities separate, see the multi-account management guide, and for how to construct session ids, see the rotating residential docs.

FAQ

How many accounts per proxy is safe?
For anything you log into, one account per dedicated IP. Login-based work does not scale beyond that, because a shared address links the identities and makes them share a rate limit. Stateless traffic has no account ratio at all; you size it by concurrent sessions instead.

How many accounts per IP can I run on the same platform?
On the same platform, plan for one account per IP for any login-based work. Stacking accounts behind one address tells that platform the identities are related and forces them to compete for the same per-address rate limit. Give each account its own dedicated address.

What is a good accounts per IP ratio for residential proxies?
For login-based use the ratio is one account per address, and a sticky residential session is best-effort and TTL-bound rather than truly fixed. For stateless work there is no account ratio: residential proxies are sized by concurrent sessions and billed per GB, not by accounts.

Do I need a dedicated IP for every account?
For persistent identities, yes. A social profile, seller account, ad account or mailbox is the same identity each login, so it should present the same address each time. A static ISP address gives you that fixed address on a 30-day lease with unlimited extensions and unlimited bandwidth.

Can I run multiple accounts on one residential IP?
Not for accounts you log into. A rotating pool changes the address, which breaks a live session, and any shared address links the identities behind it. Use one static ISP address per persistent account, and keep residential rotation for stateless requests that hold no login.

Does a higher proxy to account ratio save money?
For logins, no: sharing addresses across accounts links identities and trips rate limits, and the real cost is broken separation, not the proxy line item. Savings come from matching the model to the work, per IP for persistent identities and per GB for stateless volume.

Start with the 256MB free trial and size a small fleet against your real use case before you scale.


Leave a Reply

Your email address will not be published. Required fields are marked *