1,000 publishers. One API key.No BD calls.
Skip 18 months of partnership negotiations. Skip the firecrawl-and-pray approach. Drop xpay✦ into your retriever and ship Tuesday with licensed content from 1,000+ publishers across 16 verticals.
The current options all suck.
Scrape and pray. Negotiate for 18 months. Or buy stale Common Crawl. Pick your poison.
Scraping breaks every Tuesday
Cloudflare changes a fingerprint, your firecrawl pipeline 403s, your RAG eval drops. You burn an engineer's week debugging anti-bot detection.
Common Crawl is months stale
Your customers ask about last week's product launch. Your RAG returns last quarter's rumor. Your retrieval evals look fine, but users churn.
Licensing deals don't scale to startups
You can't sign a $1M/year deal with WSJ. You can't even get a meeting. Meanwhile your competitors quietly scrape and pray no one sues.
xpay is the fourth option.
Skip the lawsuits. Skip the licensing deals. Skip the attribution headaches. Pay-per-query, one account, every publisher in our network.
Three-line drop-in retriever
Native SDKs for LangChain, LlamaIndex, and CrewAI. Replace your existing web search tool with xpay in under five minutes.
Pay-per-query, no minimums
Sub-cent queries on most content. Spend $5 to evaluate, $50 to launch, scale from there. No BD call, no MSA, no procurement.
Licensed = enterprise-ready
When your indie RAG SaaS lands a Fortune 500 customer, your sourcing layer is already audit-ready. Receipts and provenance are built in.
Three lines. Ship Tuesday.
Native SDKs across the agent stack. Drop in xpay where you already retrieve.
from langchain.tools import XpayRetriever
from x402 import Wallet
retriever = XpayRetriever(
wallet=Wallet.from_env(),
max_cost_per_query=0.10,
)
# Drop into any chain
docs = retriever.get_relevant_documents("AI agent regulation 2026")
# docs are licensed, fresh, with publisher attributionLive in five minutes.
01
Get an API key
Sign up at app.xpay.sh, fund a wallet (USDC on Base), get your API key. New accounts ship with $50 in free credits.
02
Drop in the retriever
Three lines of code. Native SDKs for LangChain, LlamaIndex, CrewAI, Stagehand, Playwright. Or hit the REST API directly with x402 headers.
03
Get licensed content + receipts
Each query returns the document, publisher attribution, and a settlement receipt. Pay only for what you actually retrieve. Cap spend per project, per user, per domain.
Built for agents like these
RAG SaaS platforms
Vectara, Ragie, LlamaIndex CloudInternal enterprise RAG
Built on LangChain or CrewAIVertical AI startups
YC AI batch, indie RAG productsStop scraping.Start retrieving, legally.
One API key. 1,000+ publishers. Per-query pricing. Receipts and attribution by default.
Self-serve · $50 free credits · No BD calls · No MSA requiredRAG Builders — FAQ
Common questions from agent builders evaluating xpay
Those are scraping infrastructure — they help you fetch the page, but the legal status of the content you pulled is the same gray zone. xpay is the licensing layer: every retrieved document is paid-for and citable.
Per-query pricing ranges $0.001-$0.50 depending on publisher and content. For most RAG workloads, this comes out cheaper than maintaining a scraping stack (proxies, browser fleets, anti-bot retries) and far cheaper than a licensing deal.
Yes. xpay is the retrieval source, not the vector store. Index the documents you retrieve into Pinecone, Weaviate, pgvector — whatever you already use.
Yes. Set spend caps at the wallet, project, or per-query level. Built-in rate limits and budget enforcement.
New accounts get $50 in API credits. Enough to evaluate, build a prototype, and see real retrieval quality.