Read the web in LangChain through ProxyHat residential proxies — a document loader and an agent tool with rotating IPs, geo-targeting, and sticky sessions.
Tip
Recommended proxies — ProxyHat residential IPs. Every feature in this package is tested end-to-end against ProxyHat and works great. First-class integration; also works with any proxy, or none.
LLM pipelines that read the open web from datacenter IPs get blocked, rate-limited, and served bot-detection pages instead of content. This package routes LangChain's web reads through ProxyHat's residential IPs (50M+ across 148+ countries) — a fresh IP per page by default, or one pinned IP for a whole multi-page load. Two first-class LangChain components:
ProxyHatLoader— aBaseLoaderthat fetches URLs and yieldsDocumentobjects for RAG, indexing, or summarization.ProxyHatFetchTool— aBaseToolthat lets an agent fetch a page's text from a residential IP.
pip install langchain-proxyhatlangchain-core is a peer — bring your own version (>=0.3), the same one your app and other LangChain integrations already use.
Load pages as Document objects:
from langchain_proxyhat import ProxyHatLoader
# An API key auto-selects an active residential sub-user:
docs = ProxyHatLoader(
["https://example.com", "https://example.org"],
api_key="ph_your_api_key",
country="us",
).load()
for doc in docs:
print(doc.metadata["source"], doc.metadata["status_code"], len(doc.page_content))Give an agent a residential-IP fetch tool:
from langchain_proxyhat import ProxyHatFetchTool
tool = ProxyHatFetchTool(api_key="ph_your_api_key", country="us")
# Use directly...
html = tool.invoke({"url": "https://example.com"})
# ...or hand it to an agent alongside your other tools.
# from langchain.agents import create_agent
# agent = create_agent(model, tools=[tool])Get an API key at proxyhat.com.
Pass them explicitly or via environment variables — options win over env:
| Option | Env var | Notes |
|---|---|---|
api_key |
PROXYHAT_API_KEY |
Auto-selects an active sub-user with remaining traffic |
sub_user |
PROXYHAT_SUBUSER |
Pick a specific sub-user by uuid or name (with an API key) |
username |
PROXYHAT_USERNAME |
Explicit gateway proxy_username (skips the API) |
password |
PROXYHAT_PASSWORD |
Explicit gateway proxy_password |
The same knobs are available on both the loader and the tool:
ProxyHatLoader(
["https://example.com"],
api_key="ph_your_api_key",
country="us", # ISO code or "any" (default)
region="california",
city="new_york",
filter="high", # AI IP-quality tier
sticky="30m", # keep one IP across the whole load; omit for rotating
protocol="http", # or "socks5"
)By default the loader is rotating: each page is fetched over a new connection with a stable gateway username, so ProxyHat hands out a fresh residential IP per page. Set sticky (e.g. sticky="30m" or sticky=True) to pin one residential IP across every page in the load — the right choice for paginated sites, logged-in sessions, or anything that must look like a single user. On ProxyHatFetchTool, sticky keeps the same IP across the agent's successive calls.
Both components resolve your gateway credentials once (via the official proxyhat SDK — an API key auto-picks an active sub-user, or you supply username/password directly), then build a ProxyHat gateway connection URL with the SDK's targeting grammar and fetch through it with httpx. Rotating loads reuse a stable username so the gateway rotates the IP per connection; sticky loads reuse one session id so every request exits from the same IP.
Publishing to the LangChain integrations registry is a planned follow-up.
MIT © ProxyHat