Licensed source/43-state launch design/Flat subscription
The data business with no scraper.
Scrapeless Data is building a licensed US consumer and business records API. We license and index already-compiled, provenance-known data under supply contracts, and will sell records by API on a flat subscription: one call per record, a provenance object on every response, and no scraping anywhere in the pipeline.
{ "record_id": "rec_01HZX4M7P0", "name": "Jane Example", "city": "Cheyenne", "state": "WY", "suppressed": false, "provenance": { "supply_contract": "sc_2026_wholesale_0417", "source_id": "src_compiled_us_b2c_v9", "contracted_at": "2026-07-14" }, "latency_ms": 41 }
The thesis
Everyone who builds software needs a record. Nobody sells them one cleanly.
The record market is not short on data and not short on developer APIs. It is short on records you can show where they came from. In 2026 the FTC warned thirteen brokers with penalties up to $53,088 per violation, and a record with no paper trail is a liability the buyer inherits. Scrapeless Data starts from the opposite end: the source is contracted first, and every response names its origin.
How it works
One call per record. Provenance attached.
The API design is deliberately small. You send one GET request with the identifier you hold, such as an email address. You get back one record with a provenance object naming the supply contract, the source file, and the contract date. Suppressed records return as suppressed instead of silently vanishing, and latency is reported on every response so agents can budget against it.
The protocol surface
Readable by people. Discoverable by agents.
The machine surface is live today and truthful today. The UCP manifest declares what is planned, the ACP configuration declares accepting_orders false with every catalog entry flagged honestly, and the AP2 trust configuration publishes a real signing key while mandate issuance waits for activation. An agent that reads this surface learns exactly what exists, what does not yet, and when that changes.
Universal Commerce Protocol
The agent discovery layer: how AI agents find what an organization offers.
Agentic Commerce Protocol
The checkout execution layer: how AI agents place and confirm orders.
Agent Payments Protocol
The cryptographic trust layer: signed, bounded, expiring transaction authorizations.
Check any domain for the same surface with the free endpoint /api/check-endpoint?domain=example.com, or read the docs and the glossary.
Built for agents
Your agent can register without you.
Most software is now written by describing it, and a 2026 arXiv qualitative study found 59 percent of experienced engineers ship AI-generated code they do not fully understand. That is who this API serves, so registration itself is machine-first: an AI agent can create a pending account through a documented JSON endpoint at /agents, no human in the loop, and be onboarded in launch order like everyone else.
Questions, answered plainly
What is Scrapeless Data?
Scrapeless Data is an operating company building a licensed US consumer and business records API. The method is the whole thesis: we license and index already-compiled, provenance-known data under supply contracts, and will sell records by API on a flat subscription. We do not scrape, and every planned response carries a provenance object naming the supply contract, the source, and the contract date.
Can I look up records today?
Not yet. You can create an account today, and accounts are onboarded in launch order, but API keys are issued at activation and no records are served before then. Every response shown on this site is labeled illustrative, the ACP catalog carries truthful availability flags, and no endpoint on this domain accepts an order.
What will it cost?
Three flat subscription tiers are planned: Starter at $19.99 a month sized for 1,500 record lookups, Pro at $149 a month sized for 18,000 lookups, and Enterprise from $5,000 a month for a million or more. All numbers are planned launch pricing, subject to change. There are no credit packs, no per-seat pricing, and no metering surprises: one API call returns one record.
What is in the provenance object?
Every planned record response carries a provenance object with three fields: the supply contract id the record was sourced under, the source id of the compiled file it came from, and the date that contract was signed. That paper trail is the product. A record you cannot trace is a liability the buyer inherits, and we designed the API so no record ever ships without its origin attached.
How do AI agents work with Scrapeless Data?
The whole discovery surface is machine-readable today: a UCP manifest at /.well-known/ucp, an ACP configuration at /.well-known/acp/config.json with truthful availability flags, an AP2 trust configuration with a published signing key, plus llms.txt and ai-context.json. A machine-first registration surface for agents is documented at /agents, and accounts onboard in launch order.
Create your account.
Accounts are onboarded in launch order; API keys are issued at activation. Registration is open to people and to agents, the docs are public, and the protocol surface is live today, so whatever is building on your behalf can read exactly what happens next.