Every Tensor Proxies package includes unlimited traffic and unlimited bandwidth. Pricing is attached to the proxy itself, not to the gigabytes you push through it, so a proxy that moves 500 GB in a month costs exactly what a proxy that moves 500 MB costs.
That single pricing decision is what makes budgets predictable: you size a pool once, and the invoice stops moving. There are no bandwidth meters, no overage charges, and no per-gigabyte add-ons on any package.
Yes. Unlimited traffic and bandwidth apply to the Datacenter, Static Residential ISP, and Rotating Residential ISP packages alike. Nothing in the purchase flow meters data, and there is no data allowance to top up mid-month.
Static Residential ISP proxies are $15 per 25 proxies, which is $0.60 per residential IP with unlimited traffic included, flat at every quantity. Datacenter proxies start at $8 per 25 ($0.32 each) and step down by volume to $0.14 per proxy at 5,000+. Either way the discount ladder is attached to proxy count, because there is no data component to meter.
So a 25-proxy residential pool is $15 whether you use it for a handful of daily checks or a continuous scraping job. The only lever that changes your bill is how many IPs you hold.
Metered plans charge for the one thing data collection produces most of: bytes. That makes cost a function of how well your extraction is tuned — a page that ships heavy images or an unfiltered crawl can multiply a bill without changing a line of business logic. Per-proxy pricing decouples the two, so engineering choices stop having surprise billing consequences.
It also removes a whole class of defensive engineering. There is no reason to strip assets purely to save bandwidth, no reason to cap concurrency to protect a data allowance, and no reason to reconcile usage dashboards at month end.
There is no bandwidth meter and no per-account data cap, and pricing is not adjusted retroactively based on usage. Use remains subject to the Terms of Service, which cover abuse and prohibited activity rather than volume, so ordinary heavy traffic is exactly the case the pricing model is built for.
Read the pricing page for the word that qualifies the promise. Unlimited with a fair-use clause that throttles above a threshold is a data cap with extra steps; unlimited on a plan whose base price scales with a bandwidth tier is metering by another name. What you want is a price attached to the proxy count and nothing else, which is what per-IP pricing means here: $0.60 per static residential ISP proxy, $0.32 per datacenter proxy, no data component in the calculation.
Second thing to check is whether the residential IPs are dedicated. Unlimited traffic on an IP shared with strangers is a weaker offer than it looks, because their usage shapes how your IP is treated by targets.
Size on concurrency and per-IP request rate instead. Count the parallel sessions you need, then divide by how many requests each target tolerates from a single IP before it rate-limits. That number, not your data volume, sets how many proxies to buy — and it is adjustable at any time by ordering more.
For unlimited-traffic residential IPs, the Static Residential ISP package is the direct fit: dedicated consumer-ISP addresses at $0.60 each, no data meter, HTTP and SOCKS5 on the same credentials.
If the workload is bulk collection on targets that accept hosting ranges, the Datacenter package carries the same unlimited bandwidth at $0.32 per proxy, or $0.14 at 5,000+.
No. There is no data allowance to exceed, so there is nothing to bill as overage. The price you see per 25 proxies is the price regardless of throughput.
No — they are different things. Unlimited bandwidth means no data cap or metering. Throughput on any single connection still depends on the route, the target server, and your own concurrency.
Yes. Unlimited traffic and bandwidth apply across the full lineup, including the $8 per 25 Datacenter package.
No. Choose a quantity, allocate it across countries, and check out. Nothing about the order depends on a forecasted data volume.