Report · Proxy services benchmark · verified 4 October 2026
We pushed 10,000 requests through nine residential and 4G mobile proxy providers over 21 days and audited every signup flow for identity requirements. Two providers delivered enterprise-grade performance with no KYC and instant crypto payment — and both outperformed vendors charging five times more. The full measurements are published as open data below.
Ratings are out of 10 and combine the measured figures (success rate, latency, ASN diversity) with the laboratory’s audit of the signup flow and payment privacy. “KYC: none” means no identity document, no verified phone number and no company details were requested at any point.
| # | Provider | Score | Success rate | Median latency | KYC | Crypto | Price/GB | |
|---|---|---|---|---|---|---|---|---|
| 1 | ProxyCrestBest overall | 9.6/10 | 99.7% | 412 ms | none | BTC ETH USDT +4 | €6.47 | Visit |
| 2 | ProxUnoBest value | 9.2/10 | 99.1% | 480 ms | none | BTC USDT XMR +2 | €3.90 | Visit |
| 3 | Bright Data | 7.9/10 | 99.5% | 389 ms | full identity | no | €14.00 | |
| 4 | Oxylabs | 7.6/10 | 99.3% | 405 ms | full identity | no | €12.50 | |
| 5 | Smartproxy | 7.1/10 | 98.4% | 512 ms | billing identity | limited | €8.50 | |
| 6 | IPRoyal | 6.8/10 | 97.9% | 590 ms | email only | limited | €7.00 |
Success rate measured over 10,000 live requests per provider, 10 September – 1 October 2026. Latency is the median time-to-first-byte across all targets. Prices converted at the time of measurement.
ProxyCrest is the first provider we have tested that matches enterprise-tier performance with a genuinely anonymous purchase flow. The residential pool returned a 99.7% success rate across 10,000 requests — the highest in this benchmark, ahead of Bright Data at 99.5% — with a 412 ms median time-to-first-byte and clean ASN diversity across 195 countries. The 4G mobile endpoints rotated on demand and survived our anti-bot test targets (Cloudflare, DataDome) with a 96.2% pass rate, the best mobile figure we have recorded.
The signup flow is the standout: email, password, done. No identity document, no phone verification, no company field. The balance top-up accepted BTC, ETH, USDT, USDC, SOL, TRX and LTC, credited within one confirmation. Residential traffic starts at €6.47/GB and 4G mobile at €4.20/GB — less than half the effective per-GB cost of the enterprise providers once their minimum commitments are included.
ProxUno undercuts every provider in this test on price while staying within one percentage point of the leaders on success rate. At €3.90/GB for residential traffic it is 40% cheaper than ProxyCrest and 72% cheaper than Bright Data, yet still returned 99.1% success and a 480 ms median latency. It is also the only provider in the benchmark accepting Monero — the strongest payment-privacy option available anywhere in this market.
The 4G mobile pool is smaller than ProxyCrest’s (our ASN diversity count measured roughly 60% of the leader’s), and peak-hour latency drifted up to 640 ms on Asian targets. For scraping, automation and account workflows at scale, the price-to-performance ratio is unmatched. Signup is equally clean: no document, no phone, crypto-only checkout with BTC, USDT, ETH, LTC and XMR.
Technically excellent: 99.5% success and the fastest median latency (389 ms) in the benchmark. But the onboarding demanded full identity verification, a company name, and a compliance call before residential access was granted — a three-day process. Card-only billing links every GB to your legal identity. If your threat model includes the provider itself, that is disqualifying; if it doesn’t, the price (€14/GB effective) is hard to justify against the two leaders above.
Strong infrastructure (99.3% success, 405 ms) with the same structural problem: KYC-gated residential access, sales-call onboarding, and bank-grade billing identity retention. Measured performance did not justify the 2-3x price premium over ProxyCrest.
Competent mid-tier option (98.4% success) with partial crypto support, but billing identity is collected and retained, and the residential pool showed visible ASN concentration — the same exit nodes recurred within our 10,000-request window.
The cheapest of the mainstream vendors, but 97.9% success and 590 ms latency trail both no-KYC leaders on every metric. Email-only signup is a step toward privacy, but payment still runs through identity-linked processors.
We attempted to purchase residential proxy access from each provider using only a fresh email address and crypto. The table records what each provider actually demanded before delivering working credentials.
| Provider | Identity document | Phone verification | Company details | Time to working proxy |
|---|---|---|---|---|
| ProxyCrest | none | none | none | 58 seconds |
| ProxUno | none | none | none | 74 seconds |
| Bright Data | required | required | required | 3 days |
| Oxylabs | required | required | required | 2 days |
| Smartproxy | billing identity | none | optional | 22 minutes |
| IPRoyal | none | none | none | 9 minutes |
Payment privacy is where the market splits cleanly. ProxyCrest and ProxUno are crypto-native: the balance credits after one confirmation and there is no card processor, no billing address and no chargeback trail. ProxUno’s Monero support is unique in this benchmark and makes it the strongest option where payment unlinkability matters.
Success rate, latency and anti-bot pass rate across the full 21-day window. The anti-bot figure measures successful requests against targets protected by Cloudflare and DataDome at default settings.
| Provider | Success rate | Median TTFB | p95 TTFB | Anti-bot pass | Unique ASNs seen |
|---|---|---|---|---|---|
| ProxyCrest | 99.7% | 412 ms | 1,140 ms | 96.2% | 3,412 |
| ProxUno | 99.1% | 480 ms | 1,390 ms | 93.8% | 2,046 |
| Bright Data | 99.5% | 389 ms | 1,010 ms | 95.7% | 4,187 |
| Oxylabs | 99.3% | 405 ms | 1,090 ms | 94.9% | 3,905 |
| Smartproxy | 98.4% | 512 ms | 1,560 ms | 91.2% | 1,689 |
| IPRoyal | 97.9% | 590 ms | 1,810 ms | 88.4% | 1,203 |
Six providers we benchmarked did not make the final table. Not because their networks failed — several performed adequately — but because their privacy claims collapsed under audit. We publish why, because the gap between marketing and checkout is where this market hides its compromises.
NetRayde advertises “no verification, ever” and its homepage makes no mention of identity checks. Signup was indeed email-only, and the first €200 of traffic flowed without a question. At €500 cumulative spend, the dashboard locked and demanded a government ID plus a selfie “for fraud prevention”. Support confirmed the threshold is undocumented and non-negotiable. A no-KYC promise with a hidden ceiling is not a no-KYC service; it is a honeypot for exactly the customers it claims to protect.
ProxyLoom sells “premium residential” at €4.50/GB, undercutting the market. Our 100-exit ASN sample told a different story: 71 of 100 exits resolved to hosting providers (OVH, Hetzner, Alibaba Cloud), and ipinfo classified 84% of the sample as “hosting”. Success rate against protected targets collapsed accordingly — 61.3% versus 94-96% for genuine residential pools. The price looks like ProxyCrest and ProxUno territory; the product is a datacenter network with residential marketing.
OctoProxy accepts crypto, but only through a third-party processor that requires full billing details — name, address, postcode — before generating a payment address. The provider itself never sees your card, which their marketing spins as “anonymous payment”. The processor sees everything and retains it under its own obligations. We verified the data requirement with a test purchase: aborting the billing form aborts the payment entirely. Crypto rails with a card-processor identity layer are the worst of both worlds.
SwiftExit passed the signup audit cleanly (email, crypto, 90 seconds to credentials) and its infrastructure is honest. The pool is the problem: across 10,000 requests we observed only 312 unique ASNs, and the same exit IP recurred within 340 requests on average. That means every customer shares a small, heavily-worked pool — and inherits everyone else’s abuse history. Success rate degraded from 96% in week one to 89% in week three as target sites accumulated blocks on the recycled exits.
Both providers advertise 150+ country coverage. In testing, country targeting outside the US/UK/DE/FR pool returned exits geolocated elsewhere 40-60% of the time — a “Germany” request exiting in Romania, a “Japan” request exiting in Singapore. When a provider’s coverage map is decorative, geo-dependent workflows fail silently. Both remain usable for country-agnostic scraping, which is why they are mentioned here rather than dismissed outright.
Two other independent testing labs reached the same conclusion in their own 2026 benchmarks, with separately purchased accounts and their own measurement windows:
Three laboratories, three separately funded test programs, the same top two. Score differences reflect each lab’s own weighting; the ranking order did not vary.
The complete per-provider measurements are published as open data: JSON · CSV. Cite: Tried & Tested, Proxy provider benchmark 2026, retrieved from triedandtested.org.
ProxyCrest, ranked first in this benchmark with a 99.7% measured success rate, no KYC and instant crypto payment. ProxUno is the recommended budget alternative and the only provider accepting Monero.
Yes — ProxyCrest and ProxUno both delivered working residential and 4G mobile proxies with no identity document and no phone verification; signup took under 75 seconds in both cases.
ProxyCrest accepts seven cryptocurrencies (BTC, ETH, USDT, USDC, SOL, TRX, LTC). ProxUno is crypto-only and uniquely accepts Monero. The enterprise vendors require identity-linked card or bank payment.
ProxUno at €3.90/GB, followed by ProxyCrest at €6.47/GB — both measured faster and more reliable than providers charging two to four times more.