Billed per Mbps, unmetered
Rotating Residential · Bandwidth
The bandwidth plan moves the meter from bytes to Mbps: 100Mbps from $32/day, unmetered, the same price whether you saturate the line or barely touch it. Exits come from the same 90M+ residential pool, with the same rotation, sticky sessions and city/ASN targeting as the per-GB plan. It fits pipelines that run continuously at steady throughput — full-site recrawls, media downloads, LLM corpus collection. Concurrency is uncapped, bandwidth is the only ceiling, and the cost is known before the job starts.
- Unmetered traffic
- No concurrency cap
- 99.9% success rate
- Port or bandwidth tiers
07
Specifications
- IP type
- Rotating residential
- Bandwidth
- Dedicated 100Mbps – 1Gbps
- Traffic cap
- Unlimited (fair use policy)
- Rotation
- Configurable / auto-rotate
- Concurrency
- Optimised for high throughput
- Threads
- Unlimited
- Protocols
- HTTP / HTTPS / SOCKS5
04
Where it fits
Continuous full-site crawls
Catalogues, listings, job boards and property data get recrawled daily at a volume that barely moves. Once the bandwidth is bought, page count stops affecting cost.
LLM corpus and training-data collection
Text, images and media pulled together, hundreds of gigabytes in a single run. Per-byte billing gets unpredictable fast; per-Mbps leaves crawl rate as the only thing to tune.
Bulk media download
Video, audio and high-resolution images burn tens of times the traffic of a text page. At that ratio, bandwidth billing is the only model that stays legible.
Always-on monitoring
Price checks, availability probes and ad verification run around the clock — dense in requests, tiny per request. A fixed line prices that far more predictably than metered traffic.
Engineering detail
Technical detail
- Bandwidth is sold per port or as a pooled total, from 100Mbps, adjustable by the day so peak capacity is not a permanent bill.
- A saturated 100Mbps line moves roughly a terabyte a day; none of it is metered and nothing is throttled after a hidden threshold.
- There is no cap on concurrent connections — bandwidth and the target's own tolerance set the practical limit.
- Bandwidth splits across sub-accounts, so crawling, downloading and monitoring never starve each other.
- Rotation, sticky sessions and city/state/ASN targeting use the same configuration as the per-GB plan, so switching over needs no code change.
Coverage
One network behind all four products — changing the billing model does not change how you connect or how you target.
All 33 markets- 195 countries & regions
- Available
- All 50 US states
- Available
- City-level targeting
- Precise
- ISP & ASN targeting
- Supported
FAQ
Questions about this product
Is an unlimited bandwidth proxy really unlimited?
The bytes are genuinely unlimited; the ceiling is the line you bought. Saturating 100Mbps for a full day moves about a terabyte, none of it metered and none of it throttled past some undisclosed threshold. The only real limits are the line rate itself and the uses ruled out in the acceptable use policy.
Unlimited bandwidth vs pay-per-GB residential proxy — which is cheaper?
Work out your average throughput first. If the job runs most hours of the day at a steady rate, bandwidth wins, and roughly a terabyte a month is the crossover point. If it runs twice a week for a few dozen gigabytes, per-GB wins because idle time is free. When it is genuinely unclear, run a week on the per-GB plan and read the usage curve in the console.
Is there a concurrent connection limit?
Not on our side. A thousand connections or ten thousand cost the same; what decides throughput is the bandwidth you bought and how hard the target will let you push. In practice, raise concurrency until the line sits near 70% utilisation, then tune against the target's error rate rather than ours.
Can I use residential proxies to collect LLM training data?
Yes, and it is one of the most common uses of this plan — a multi-terabyte mixed-content crawl is hard to budget per byte and straightforward to budget per Mbps. Collection has to stay on publicly accessible content and respect each site's robots directives and terms of service. That responsibility sits with the customer, and the acceptable use policy says so plainly.
Get one request working. Scale after that.
Test before you commit, with an engineer on your integration.