How we test and rate proxies
Our scores come from running proxies against real targets, not from press releases. Here's exactly what we measure, how we label confidence, and why you can trust the verdict.
What we score
Every provider gets an overall rating built from four sub-scores, each out of five:
- Success rate — does it actually get the page back on hard targets? This carries the most weight.
- Speed — latency and throughput under real concurrent load.
- Value — cost per useful result, not just headline price per GB.
- Ease of use — dashboard, docs, endpoints, and how fast you're productive.
Tested vs researched
We label every provider so you know how much weight to give the score:
Two confidence levels
Tested — we routed production scraping jobs through it and measured the results first-hand. These verdicts carry our full confidence.
Researched — compiled from the provider's public documentation, current pricing, and credible community reports. These are honest editorial assessments, clearly marked as not-yet-benchmarked. We upgrade them to Tested as providers enter our pipeline.
How we actually test
Our test bed is our own work: large-scale scraping of stubborn targets like search engines and review platforms. The setup that produces our numbers:
- Real jobs against real anti-bot defenses — not synthetic ping tests.
- Residential, ISP, and datacenter pools compared on the same targets.
- Camoufox (hardened Firefox) under xvfb, with consent-wall handling and token-replay deep pagination, so the browser layer is consistent across providers.
- Per-outlet IP rotation and region-specific ports, so geo-accuracy is part of the test.
A real lesson from the bench
Methodology is only as good as what it surfaces. A concrete example: on hard review pages, datacenter proxies dropped to roughly a fifth success rate while residential held up. That single finding is why our scoring weights success rate above speed — a fast proxy that gets blocked is worthless, and only first-hand testing makes that visible.
Pricing and freshness
Pricing is the lowest published rate at the time of writing, always shown with an "as of" date and source. Proxy pricing changes often — we note when figures were captured and tell you to confirm on the provider's site before buying.
Independence
ProxyPeers is reader-supported through affiliate links, which may earn us a commission at no cost to you. Earnings never influence rankings — we routinely recommend free and low-cost options that pay us little. The score reflects what works for the job, full stop.