data studies

ProxyVero measured our residential pool at 89.4% success and 60% on Google. We downloaded its full 50-provider dataset to find out whether 60% is bad.

We Pointed Free Proxies at 21 AI Sites: Which Edges Block Them, Which Wave Them Through, and the Four Kinds of Wall

A measured study, 24 August 2026: three runs of 50 free proxies from our own list, each sent to a dozen AI sites. Only 20 to 27 of 50 could carry HTTPS at all. claude.ai challenged 18 of 27, Character.AI refused half, Poe challenged a third, while DeepSeek, Civitai, Janitor AI, Mistral and x.ai let every one through, Hugging Face answered three with a 429, and Google Flow showed four a country page. The numbers, the four different walls behind them, and what they mean if you rely on free proxies.

Scraping API Costs Run 50x Apart, and the Plan You Buy Is Not Why

Four scraping APIs, converted to one number: dollars per 1,000 requests. The range runs from $0.0749 to $3.75, and on one vendor both ends of a 25x spread sit inside the same $30 plan.

Correcting Our Own Price Data: Four Errors We Found Re-Checking 20 Vendors

We re-verified all 20 residential proxy vendors we track against their own pricing pages on 10 August 2026, one week after the first capture. Four of our records were wrong, and one of them was wrong in the direction that suited us.

The Permanent Sale: 8 of 20 Proxy Vendors Are Discounting and Only 1 Says When It Ends

We re-read the residential pricing page of 20 proxy vendors on 10 August 2026. Eight were running a discount. One of the eight published an end date.

Headless Browsers Are Not Free: What Puppeteer Costs in Proxy Bandwidth

A headless browser downloads the whole page. Behind a residential proxy that is billed by the gigabyte, and at 6.54 MB per protected page fetch the meter runs faster than most teams budget for.

We Checked 580,000 Free Proxies 110 Million Times. Two Thirds Never Worked Once.

Our engine has verified 580,000 free proxies 110 million times across 110 source lists. Most never worked, the best source has a 40% hit rate, the worst 0.19%.

Residential Proxy Prices in 2026: What 16 Providers Advertise vs What They Charge

We checked what 16 residential proxy providers advertise against what they charge a new customer with no commitment. The median gap is 2.5x. The widest is 20x.

The Other End of a Residential Proxy: What the Person Supplying It Earns

Residential proxies are resold consumer bandwidth. We checked what the apps buying that bandwidth pay per GB, against what providers charge for it.

Scraping API Pricing: Why 250,000 Credits Is Not 250,000 Requests

Scraping APIs bill in credits, and one request can cost 1 credit or 75. We read the credit tables at ScrapingBee, ZenRows and ScraperAPI and did the arithmetic.