x402 Entity-ID Resolver

Name in, official IDs out.

Give it one ambiguous name or ticker — "Apple", "ETH", "Stripe" — and get back ranked entities carrying verified canonical IDs from every official registry, the right one ranked first. The exact cross-registry lookup an LLM guesses — and gets wrong.

$0.005 / call USDC on Base gasless · EIP-3009 no account · no API key
GET/resolve?q=apple
company Apple Inc. confidence 0.98
✓ sec✓ gleif✓ wikidata
SEC CIK0000320193
TickerAAPL
GLEIF LEIHWUPKR0MPOU8FGXBT394
Wikidata QIDQ312

Every ID is pulled live from its own authoritative registry and tagged with the sources that agreed — paste the CIK into sec.gov or the LEI into gleif.org and check it yourself.

Why this exists

Anyone who has wired an LLM into a financial or data workflow has hit the same wall — and a steady stream of engineering write-ups, papers and post-mortems describe it: language models confidently invent identifiers. Ask for Apple's SEC CIK or a counterparty's LEI and you get something format-perfect and wrong. Entity resolution — mapping a messy name to its one canonical record — is one of the oldest unsolved chores in data engineering, and it became urgent the moment agents started acting on the answers. This service was built for exactly that gap: it never recalls an ID, it looks each one up live in the entity's own official registry and tells you which sources agreed.

Why you can trust it

Straight from the registry

Every ID comes from SEC EDGAR, GLEIF (the official global LEI issuer), Wikidata and CoinGecko. We don't invent, store or own any of it.

Checkable sources

Each result tags which sources agreed, so you can verify any ID at its own registry yourself.

Deterministic

Same name → same verified IDs, every call. No model drift, no hallucinated tickers.

Stateless

Pay-per-call. No account, no API key. Queries aren't stored beyond transient operational logs.

What it's for

KYC / onboarding

"Stripe" → the exact legal entity + its GLEIF LEI before you onboard a counterparty.

Compliance reporting

A counterparty name → its LEI, which many filings now mandate.

Ticker disambiguation

"ETH" → Ethereum ranked first, with any same-ticker ETF or trust kept below it. The canonical asset wins, not a look-alike fund.

Data pipelines

A name → CoinGecko slug to wire a price feed, or → CIK to pull SEC filings.

Agent tool-use

Hand an LLM the verified canonical key it can't recall reliably, so it can call the next API.

Endpoints

Try it now — free

curl "https://entityresolver.xyz/demo?q=apple"

That is the real resolver, no payment needed, capped per IP per day. When you're ready for production, switch to GET /resolve?q=: an unpaid request returns HTTP 402 with payment terms, and a wallet holding USDC on Base signs an EIP-3009 transfer inline (gasless for the buyer); the facilitator verifies + settles. Agents: see /llms.txt or /openapi.json.

Use from Claude or any MCP agent

{
  "mcpServers": {
    "entityresolver": {
      "command": "npx",
      "args": ["-y", "entityresolver-mcp"],
      "env": { "BUYER_PRIVATE_KEY": "0x…" }
    }
  }
}

The entityresolver-mcp package exposes two tools: demo_resolve (free, works with zero config — omit the env block) and resolve_entity (production, pays $0.005 USDC per call from a dedicated low-value wallet you fund with a few cents of USDC on Base; gasless, no ETH needed).