IPRoyal vs Smartproxy (Decodo): Residential Proxy Guide

For rotating residential proxies, IPRoyal and Smartproxy—now branded as Decodo—publish meaningfully different commercial and session models. IPRoyal is the clearer shortlist for irregular traffic that may sit unused and for flows that request a sticky session longer than 24 hours. Decodo is the clearer shortlist when a new buyer wants a clearly published small trial, lower displayed plan subtotals at common self-service tiers, or US ZIP targeting or ASN targeting. Those are product-fit observations, not performance verdicts, and Decodo documents restrictions on combining ASN with other fine geo controls.

Do not decide from a claimed pool total or generic success percentage. A pool headline does not reveal how many unique exits are live in the required market, how often they repeat, whether the requested targeting is fulfilled, or whether the destination returns the data the workflow needs.

The Databay Research Team reviewed IPRoyal's public residential pricing, residential quick-start guide, and help-center billing pages, plus Decodo's Smartproxy rebrand notice, residential pricing, location FAQ, and public endpoint documentation on August 13, 2026.

This is desk research. We did not create either account, submit payment, inspect a private dashboard, contact support, route traffic through either network, or test a refund. Prices, availability, feature labels, and terms can change after publication. Follow the source links and capture the plan, cart, renewal, tax, validity, cancellation, refund, and acceptable-use terms shown to your own account before paying.

Databay sells residential proxies and competes with both providers. That gives us a commercial bias. To make the analysis auditable, this guide links to the providers' pages, labels their performance statements as claims, publishes no star rating, and supplies a test that can reject IPRoyal, Decodo, or Databay.

Decodo's rebrand page says Smartproxy officially became Decodo on April 22, 2025. It describes the business as the same company with the same team and infrastructure, says accounts and subscriptions moved automatically, and states that existing Smartproxy endpoints remain backward compatible.

The same notice identifies decodo.com and decodo.cn as official domains and warns that smartproxy.org and smartproxy.cn are impersonators. Buyers arriving through the retired name should verify the domain before entering credentials or payment details.

That means “IPRoyal vs Smartproxy” is currently an IPRoyal-versus-Decodo buying decision, not a comparison with a discontinued service. A buyer may still encounter smartproxy.com tutorials or legacy gateway hostnames. Use the hostname generated in the current dashboard, record it in configuration, and follow any migration notice rather than replacing a working endpoint merely because the public brand changed.

The rebrand also creates a research trap: older Smartproxy reviews may describe a retired price, pool claim, dashboard, or product mix. Check their publication date and then verify every deciding fact against a current Decodo source.

Both services bill rotating residential traffic by transferred data, but their public entry packages and renewal models differ. The figures below are public page observations from August 13, 2026, not quotes or future price guarantees.

IPRoyal labels subscription pricing as a five-percent discount from pay as you go. Its pricing page also advertises much lower enterprise-scale rates, but the lowest headline requires a volume far above the small cards. Decodo likewise advertises a lower “starting from” rate than its small regular tiers. Compare the package at the volume you will actually buy, not the largest-volume marketing number.

At the exact public 10 GB monthly tier, Decodo's displayed plan subtotal was lower on the review date. That does not settle all-in cost. Its own pricing page also displayed conflicting pay-as-you-go figures in the table and FAQ copy, so verify the dashboard checkout rather than treating either figure as settled. IPRoyal says purchased residential traffic does not expire under standard conditions. Decodo says unused subscription traffic normally expires at renewal, downgrade, or an unchanged new cycle, but documents an upgrade exception that transfers unused traffic; cancellation leaves the balance available until subscription expiry, and unused pay-as-you-go traffic remains valid for 12 months. Tax treatment also differs by buyer. A lower purchase subtotal can lose if valid responses require more billable bytes, retries, review, or engineering time.

Use this calculation for each candidate:

all-in cost per usable output = (plan + tax + overage + compute + storage + review + support effort) / validated outputs

Keep failed and invalid tasks in the denominator appropriate to the decision. Comparing price per gigabyte while omitting challenge pages, incomplete records, location misses, and human review rewards the wrong provider.

IPRoyal's quick-start guide documents country and region selection, available state or city controls, random rotation, sticky TTL from one second to seven days, IP allowlisting, and dashboard usage reporting. Its current residential documentation also lists HTTP(S), SOCKS5, username/password authentication, and unlimited simultaneous sessions as product features.

Decodo's residential pricing page lists continent, country, state, city, ZIP, and ASN targeting, rotating and sticky sessions, HTTP(S) and SOCKS5, and credential or source-IP authentication. Its quick-start documentation qualifies those labels: ZIP targeting is US-only, the state example is US-specific, and ASN cannot be combined with city, state, continent, or ZIP. The guide also requires an explicit session value for sticky SOCKS5 configuration. Decodo's whitelisted-IP guide documents an HTTPS-only limitation for source-IP authentication. Its custom-session documentation gives a 10-minute default and a configurable duration from one minute to 24 hours. The current dashboard should remain the source for the exact hostname, port, and username parameters assigned to an account.

A maximum sticky duration is not an availability promise. IPRoyal explicitly notes that a residential participant can leave before the requested TTL. The same operational risk applies to any peer-powered rotating pool. The client needs bounded recovery for replay-safe work and must reconcile an ambiguous state before replaying a request that may have taken effect.

IPRoyal's current quick-start page publishes a 64-million-plus residential pool across more than 195 countries. Decodo's residential pricing page publishes 115-million-plus addresses in more than 195 locations, while its location FAQ displays a generic 125-million-plus IP headline. Even within one provider's public site, totals can reflect different update dates, counting windows, products, or definitions.

Do not turn those numbers into a ratio such as “Decodo has almost twice the usable network.” Public totals do not disclose the live, unique exits available to your account in one country during your test window. They can also count an address seen over time rather than simultaneous inventory. Ask each provider for its counting method, then measure observed unique exits, repeats, requested-location fulfillment, and ASN distribution in the same bounded sample.

Provider-published latency, uptime, and success rates are also not interchangeable unless the destination, request, region, client, sample window, concurrency, timeout, retry rule, validation rule, and treatment of failures are the same. A successful IP-check request does not establish that an authorized price, ad, catalog, or market-research record is complete.

The defensible statement is narrower: both providers publish large global rotating residential products and common controls; neither public marketing page can predict your target-specific usable-result rate.

Decodo publishes a 3-day, 100 MB residential trial for eligible new users and requires payment details to activate it. Its page says the selected plan activates when the trial ends unless the user cancels. Its refund page limits the described 14-day option to an eligible self-service first purchase with no free-trial use, no cryptocurrency payment, no enterprise plan, and usage no greater than 20 percent or 1 GB, whichever comes first. Read the current contract and save the exact conversion date before beginning.

IPRoyal does not publish a guaranteed standard self-service residential trial. Its residential quick-start guide tells ordinary buyers to use the one-gigabyte paid entry; its enterprise page says verified company representatives may qualify, and its Terms make any limited trial discretionary and subject to eligibility or payment-method conditions. Its residential refund guidance primarily covers a connection or service problem that remains unresolved for 24 hours; using more than 100 MB normally disqualifies the request. It separately describes accidental-renewal requests within 72 hours when newly allocated traffic remains unused. Its help center also says completed orders cannot simply be cancelled, while subscription renewal can be disabled.

IPRoyal markets purchased residential traffic as non-expiring. A current help article qualifies pay-as-you-go persistence with “standard conditions” and notes that account-specific activity requirements may apply. That is a reason to capture the rule shown in the dashboard, not a reason to assume the balance survives indefinitely in every account state.

Before testing either service, record:

1. Product, pool, package, purchase total, tax, renewal, and traffic validity.
2. Trial start, trial end, automatic conversion, cancellation path, and refund exclusions.
3. Allowed destinations and use cases, account verification, and data-handling terms.
4. Included locations, targeting controls, protocol, authentication, session rules, and limits.
5. The maximum traffic that preserves any refund eligibility.

Use the same lawful, authorized workload for IPRoyal and Decodo. The sample should be small enough to stop safely but representative enough to expose the required markets, response sizes, session lengths, and normal variance. Do not rotate around a destination refusal, CAPTCHA, account control, 403, or 429; one source-wide request budget applies across every exit.

Keep the raw counts beside every percentage and publish the time window. If one provider ran at a different time, location, request mix, or concurrency, label the comparison rather than averaging the runs together. The rotating proxy trial checklist expands this scorecard into a repeatable qualification process.

Shortlist IPRoyal first when traffic is bursty, the ability to retain unused pay-as-you-go traffic is economically important, or an authorized multi-step flow needs a requested sticky window longer than 24 hours. Confirm any account-activity rule and test whether the live exit actually survives the required flow.

Shortlist Decodo first when the 3-day qualification offer is useful, the current regular package better matches a predictable monthly volume, or US ZIP selection or ASN selection is mandatory. Verify the documented targeting combinations, trial parity, automatic conversion, tax, location fulfillment, and the 24-hour custom-session ceiling.

Shortlist both when target-specific success, live inventory, operational support, or cost per usable output matters more than the public feature difference. The providers' own claims cannot answer those questions for your destination.

Shortlist neither when an official API, partner feed, export, licensed dataset, or direct route already answers the business question with better permission, structure, and cost. A proxy changes network origin; it does not create authorization, quota, or data quality.

Databay is a third commercial option, not a neutral referee. Its residential pricing page publishes one-time traffic packages rather than a monthly subscription, with current geo and session controls on the product page. Databay does not offer a free trial, and its free proxy list is not a sample of the paid residential pool.

Apply the same scorecard to Databay. Reject it if its required market, protocol, session behavior, usable-result rate, support, validity, or all-in cost misses a predeclared gate. No provider can guarantee access to a destination, and a larger set of exit addresses does not grant permission to bypass a control.

Whichever service you test, keep the workload within the destination's rules and Databay's Acceptable Use Policy. A comparison is useful only when it produces a reproducible buy, reject, or tightly scoped retest decision.