216M+ IPs across 199 countries.
Serious routing, effortless surface. Health-gated multi-vendor selection, honor-or-fail geo pinning, and retries that rotate pools: all behind one flag on one endpoint. Metered per GB from a single credit balance; failed requests bill $0.
+193 moreWhere Spider can route from.
Click any country to see its proxy pool, supported types, and a pricing summary.
Largest pools by IP count.
Not another reseller.
Anyone can resell a pool and print an IP count. What makes this network fast is the routing layer: the decisions it makes before and after every dial. None of it needs configuration: it's the default behavior behind proxy_enabled.
Health-gated selection
The router checks upstream pool health when it picks a route: a degraded pool is skipped at selection time, not discovered dead after your request already failed. One source going down never silently absorbs your traffic.
Retries that actually rotate
Every retry lands on a genuinely different upstream pool instead of hammering the one that just failed. In production this cut retry-storm failures roughly in half for country-pinned traffic. Retries are automatic and cost nothing extra.
Honor-or-fail geo pinning
A country pin either egresses from that country or fails cleanly. It will never quietly route through the nearest region and misreport the origin: the geo you asked for is the geo you get.
Capacity-aware routing
A live bandwidth signal per upstream pool steers traffic away from sources that are out of headroom, instead of blind round-robin into rejections. Multiple vendors sit behind every pool, so exhaustion in one never empties the endpoint.
Automatic clean-IP escalation
When a target starts blocking, requests escalate to a vetted fraud-free IP pool on their own. You never manage that tier, flag it, or pay attention to it. It engages when needed and steps aside when not.
One key, every surface
The same API key drives the Crawl API (a bearer header) and the direct HTTP/HTTPS/SOCKS5 proxy (the proxy username). Same rotation, geo-targeting, and retry logic on both. Switch surfaces by changing the request shape, not the credentials.
All 199 countries.
Route through any country.
Two ways in, one API key: the Crawl API for managed end-to-end requests, or a direct HTTP/HTTPS/SOCKS5 proxy for clients you already use. Rotation and retry are the default on both. There is nothing else to set up.
curl -X POST https://api.spider.cloud/crawl \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"url": "https://example.com",
"proxy_enabled": true,
"country_code": "us",
"return_format": "json"
}'Pick the right pool.
Two pools, one selector. Set proxy to residential or isp, or leave it out and let the router decide. Behind each pool, multiple upstream sources fail over automatically if one degrades.
Residential Proxies
- Trust
- Highest
- Best for
- Stealth scraping, account management, ad verification
Route through real consumer connections: a mix of peer and edge egress alongside hosted residential pools. Residential IPs carry the trust signature of genuine home networks, so they clear the strictest bot detection. If a target starts blocking, requests escalate to a vetted clean-IP pool automatically.
ISP Proxies
- Trust
- High
- Best for
- High-volume data collection, price monitoring, SERP scraping
Data-center IPs registered to legitimate internet service providers. Faster than residential with good trust scores. Ideal for large-scale operations where speed matters more than stealth. The default pool when you enable the proxy without picking one.
Common questions.
What types of proxies does Spider offer?
Spider offers two publicly selectable proxy pools: residential and ISP. Residential proxies route through real consumer connections for the highest trust scores. ISP proxies use data-center IPs registered to internet service providers for fast, stable, high-throughput connections. Behind each pool the router aggregates multiple upstream sources (including a vetted clean-IP pool that engages automatically when a target starts blocking), so you pick a trust level, not a vendor.
How many proxy locations are available?
Spider provides proxy coverage in 199+ countries across every inhabited continent. The total pool exceeds 200 million IPs, with the largest coverage in the United States, United Kingdom, Germany, Brazil, and Japan.
Do I need to manage proxy rotation myself?
No. Spider handles automatic IP rotation on every request. You can optionally pin a session to a single IP using sticky sessions, or let the system rotate IPs automatically for maximum anonymity.
How do I target a specific country?
Pass the two-letter ISO country code in the "country_code" parameter of your API request. For example, set "country_code": "de" to route through a German proxy. You can also set "proxy" to "residential" or "isp" to choose a pool. Country pinning is honor-or-fail: the request either egresses from the country you asked for or fails cleanly. It never silently routes from somewhere else.
What is the average proxy latency?
Residential proxies average 150-300ms depending on the target country. ISP proxies average 50-120ms. Spider automatically selects the fastest available proxy in your target location.
Is there a bandwidth limit on proxy usage?
There are no hard bandwidth caps. Usage is metered: you pay per gigabyte of bandwidth plus a small compute charge per minute of CPU, billed from one prepaid credit balance. Failed requests are billed at $0, and your balance never expires. There is no subscription and nothing resets monthly.
Can I use proxies with headless browsers?
Yes. Spider's proxy network integrates directly with its headless browser infrastructure. Enable proxy_enabled alongside browser rendering for stealth browsing that combines residential IPs with real browser fingerprints.
Do you support SOCKS5 or HTTP CONNECT proxies?
Yes. Spider exposes an HTTP/HTTPS proxy on proxy.spider.cloud:8888 (HTTP) and :8889 (HTTPS), plus a SOCKS5 proxy on proxy.spider.cloud:8887. All three use the same authentication scheme: your API key as the username and any optional parameters (e.g. country_code=us&proxy=residential) as the password, so rotation, geo-targeting, and retry logic stay identical to the REST API.
How does pricing work for proxy requests?
Proxy requests bill like every other Spider request: bandwidth per gigabyte plus compute per minute of CPU, denominated in credits from one prepaid balance. Enable proxy mode by setting "proxy_enabled": true. There is no separate proxy invoice or minimum commitment, retries cost nothing extra, and failed requests are billed at $0.
What happens if a proxy request fails?
Spider retries automatically: each retry rotates to a genuinely different upstream pool rather than hammering the one that just failed, which cut retry-storm failures roughly in half in production for country-pinned traffic. Pools that are currently degraded or out of capacity are skipped at selection time, so retries land somewhere healthy. Retries cost nothing extra, and requests that ultimately fail are billed at $0.
Residential and ISP Proxies, Briefly
What are residential proxies?
Residential proxies route requests through IP addresses that internet service providers assign to real households, so your traffic looks like a regular visitor browsing from home. That matters because anti-bot systems classify traffic by IP type, and residential IPs share the same address space as legitimate users. ISP proxies are the faster middle ground: data-center machines on IPs registered to real providers, with good trust scores and lower latency.
The practical split: reach for residential when stealth matters most (strict anti-bot targets, ad verification, account work), and ISP when speed and volume matter more (price monitoring, SERP scraping, large-scale collection). Both support country-level targeting, which is what makes region-specific pricing, search results, and ad placement visible from where your users actually are.
How Spider makes them simple
Most providers hand you a list of IPs and leave rotation, retries, health checks, and geo-targeting to you. Spider runs that layer as part of the API: one parameter turns the proxy on, "country_code" pins a country, and "proxy" picks residential or ISP. Rotation, failure detection, and retries against healthy pools happen automatically, and tracker and ad requests are blocked by default so you are not paying for analytics scripts you will never use.
Billing is metered from one prepaid credit balance: bandwidth at $1/GB (measured on content before transformation) plus compute at $0.001 per CPU-minute. No monthly minimums, no separate proxy invoice, and the same balance covers crawling, browser sessions, and AI extraction.
Start routing in 30 seconds.
One API key. No proxy vendor to manage.
Free balance on signup · No card required