What is the cheapest residential proxy?

There is no single cheapest residential proxy across all purchase sizes. In the conflict-free, standalone exact tiers observed on July 25, 2026, DataImpulse has the lowest 5 GB checkout at $5, GeoNode has the lowest 10 GB and 50 GB checkouts at $7.92 and $34.65, and Evomi Core has the lowest 100 GB checkout at $49.99. DataImpulse residential pricing, retrieved 2026-07-25T16:33:08.221Z · GeoNode residential pricing, retrieved 2026-07-25T15:42:45Z · Evomi residential pricing, retrieved 2026-07-25T16:33:08.221Z

“Cheapest” needs a traffic quantity and a checkout rule. The lowest advertised rate may belong to a much larger tier. A plan with the smallest entry receipt may not cover the workload. A traffic add-on may require another paid product. A promotion may not provide a durable recurring price.

This answer uses the narrowest reproducible definition: the lowest public checkout among standalone residential rows that have no recorded source conflict and publish an exact tier at the requested quantity.

The lowest exact residential checkouts

These rows are exact tiers, not estimates produced by applying a larger tier’s rate. The provider changes at 100 GB. GeoNode’s exact 100 GB checkout is $64.80, while Evomi Core publishes a 100 GB monthly checkout of $49.99, so Evomi takes the lower receipt at that checkpoint. GeoNode residential pricing, retrieved 2026-07-25T15:42:45Z · Evomi residential pricing, retrieved 2026-07-25T16:33:08.221Z

The lower receipt does not make the products identical. Evomi’s monthly traffic expires after 30 days and does not roll over. GeoNode records that purchased traffic never expires. A buyer who can consume 100 GB inside the month can compare the Evomi receipt directly. A buyer with irregular usage must price the risk of losing unused traffic instead of treating both lifetimes as equal. Evomi bandwidth terms, retrieved 2026-07-25T16:33:08.221Z · GeoNode bandwidth terms, retrieved 2026-07-25T15:42:45Z

Why the lowest advertised rate may not win

The dataset records a ProxyMesh prepaid bandwidth add-on at $0.05 per GB for a 100 GB order, but its tier has no published checkout total and the row is explicitly an add-on. It therefore cannot displace a standalone residential checkout in this ranking. ProxyMesh prepaid bandwidth documentation, retrieved 2026-07-25T17:42:41.156Z

This distinction prevents a common false result. Multiplying the published add-on rate by the order quantity creates a plausible receipt, but it does not establish the price of the required parent service. The honest state for a standalone checkout remains not published.

Large tiers create a second trap. A provider may publish an attractive unit rate only when the buyer purchases far more traffic. The correct ranking compares the complete checkout that covers the requested amount. It never borrows the largest tier’s rate for a smaller order.

“Cheapest” is a filtered claim

The snapshot contains 203 residential rows. Of those, 44 are conflicting, 29 have no published public price, and one source was blocked. Those rows stay in market coverage, but none can establish the cheapest public checkout. The counts come from the source states in the complete observed plan table, whose field evidence was retrieved from 2026-07-25T15:31:32Z through 2026-07-26T02:15:19Z.

A conflict does not become a price by averaging the claims. A not-published row does not become free. A blocked source does not inherit a permissive value. The comparison sinks each row until a reproducible public checkout exists.

Promotions also need separate treatment. A temporary or conditional price can be valid today while remaining unsuitable for a recurring budget. The durable ranking should use the list checkout, then show the promotion with its code, eligibility, and published end date. If the end date is null, the page should say it is not published.

Cheapest checkout is not cheapest useful result

The plan dataset does not contain controlled target tests. It cannot prove that the lowest checkout will reach a buyer’s target, sustain the required session, or return the required data. The cheapest usable plan is the lowest-priced row that also passes the buyer’s own acceptance test.

That test should hold the request path constant and record useful responses, retries, transferred bytes, latency, and session behavior. If a lower-priced plan produces more retries, its checkout can still be lower while its cost per usable result is higher. That second calculation requires workload evidence, not vendor pricing fields.

Published controls matter before a benchmark too. DataImpulse records rotating and sticky modes plus country, state, city, ZIP, and ASN targeting, but its concurrent-session value is not published. GeoNode records rotating and sticky modes plus country, city, and ASN targeting, while concurrency is also not published. Evomi Core records rotating and sticky modes, a 24-hour sticky limit, and unrestricted concurrent sessions. DataImpulse residential pricing, retrieved 2026-07-25T16:33:08.221Z · GeoNode residential pricing, retrieved 2026-07-25T15:42:45Z · Evomi residential terms, retrieved 2026-07-25T16:33:08.221Z

An unpublished capacity field should stop a requirement-specific claim. It should not be interpreted as unrestricted.

The buyer answer

For a small paid entry, the 5 GB DataImpulse tier is the lowest exact standalone checkout in the measured set. GeoNode leads the exact 10 GB through 50 GB checkpoints listed above. Evomi Core leads at exactly 100 GB, with a 30-day lifetime.

Choose the checkpoint closest to measured usage, then compare the cheapest covering configuration rather than only exact tiers. Keep expiry, rollover on the provider’s page, targeting, sessions, taxes, and source state beside the receipt. Recheck the linked vendor checkout before paying because this answer is an observation from July 25, 2026, not a promise that the same ladder remains live.