Skip to main content

What is a residential proxy?

What is a residential proxy?
Lucas Giordano's avatarBy Lucas Giordano · Co-founder, Notte
Last updated
TL;DR

A residential proxy is a proxy server that exits via an IP address assigned by an ISP to a real consumer device — a phone, a router, a smart TV. From the receiving site's perspective the request looks like ordinary household traffic, which clears reputation-based bot filters that block datacenter ranges by default. The foundation of any production stealth stack.

What is a residential proxy?

The single most-cited reason scrapers and agents get blocked isn't fingerprinting or behavior — it's that the IP gives them away. Datacenter IPs from AWS / GCP / Hetzner / DigitalOcean show up in reputation databases the moment they're allocated and stay there until they're recycled. Residential proxies are the workaround: route the request through an IP an ISP assigned to a real household, so to the receiving site the request looks like a normal Comcast / Verizon / Deutsche Telekom subscriber. It's the cheapest single change you can make to look less like a bot, and the highest-leverage one.

How residential proxies are sourced

There are two operating models for residential proxy networks, with very different ethics:

  1. Compensated opt-in. A consumer installs a desktop or mobile app (typically free VPN, free game, free browser plugin) that explicitly trades unused bandwidth for the service. The user's IP becomes a proxy exit when they're online. Examples: BrightData's MobileXProxy, IPRoyal, services that disclose the trade in plain language.
  2. Bundled into apps without clear consent. Some "free" apps include an SDK that turns the device into a proxy without a clear user-facing disclosure. This category is at best ethically grey and at worst illegal in some jurisdictions; reputable providers have moved away from it.

Notte partners with Massive for ethically-sourced residential infrastructure — opt-in, disclosed, compliant. The proxy ethics matter both because they should and because anti-bot vendors increasingly flag IPs from known shady networks.

When a residential proxy is necessary

Reach for one when any of these are true:

  • The target site has anti-bot defenses (Cloudflare, Akamai, Datadome, hCaptcha Enterprise, etc.).
  • You're running parallel scraping at any non-trivial scale.
  • The site differentiates content or pricing by geography (e-commerce price testing, geo-fenced media).
  • You're hitting consumer-facing services where datacenter IPs get auto-flagged.

You can skip it for development against friendly targets, internal tools whose IP allowlist already covers your dev box, or genuinely public APIs.

Notte SDK shape

Residential proxy routing is a single flag on the session:

main.py
from notte_sdk import NotteClient

client = NotteClient()

with client.Session(proxies=True) as session:
    session.execute(type="goto", url="https://target-site.com")
    snapshot = session.observe()

Setting proxies=True routes the session through Notte's residential pool. There's no separate proxy provisioning step, no rotation logic to write, no IP-pool exhaustion to track. For parallel browser execution the platform allocates distinct exits across concurrent sessions automatically.

Residential is necessary, not sufficient

A residential IP fixes the network layer. It doesn't fix:

  • A headless Chromium fingerprint that screams "automation."
  • A perfectly-timed action sequence that no human would produce.
  • A request rate from a single IP that exceeds plausible household traffic.
  • A timezone / language / fingerprint that doesn't match the IP's geographic exit.

Anti-bot is a pipeline. Residential proxies clear stage 1; they don't help with stages 2–5. The full stealth stack is residential IP + coherent fingerprint + plausible behavior + reasonable rate limits, all internally consistent.

Common pitfalls

  • Buying the cheapest residential pool you can find. Cheap pools are often heavily-used (already flagged by major sites) or sourced from grey-market apps (legally risky). The reputation of the pool matters more than the per-GB price.
  • Rotating IPs on every request. A real household IP doesn't change between page loads. Sticky sessions (one IP per session) is more plausible than per-request rotation for most workflows.
  • Mixing residential IP with datacenter fingerprint. Coherence is what gets you through; the IP and the fingerprint should agree on geography, time zone, and device class.
  • Treating residential IPs as anonymizing. They aren't private to you; the underlying device's owner can see the connection. Ethical sourcing matters.

Key takeaways

  • A residential proxy routes traffic through an ISP-assigned consumer IP, clearing reputation-based bot filters that block datacenter ranges.
  • Pool ethics matter — both for compliance and because anti-bot vendors increasingly flag IPs from shady networks. Notte uses ethically-sourced infrastructure via Massive.
  • One flag on the Notte session: client.Session(proxies=True) — no provisioning, no rotation logic.
  • Residential is necessary but not sufficient: pair with coherent fingerprint and plausible behavior to clear the full anti-bot pipeline.

Build your AI agent on the open web with Notte

Cloud browsers, agent identities, and the Anything API — everything you need to ship reliable browser agents in production.