CC
ClaimCell
IP data over MCP Launching soon

Patent and trademark data for any MCP-compatible AI tool.

Add ClaimCell to Claude, Cursor, Copilot, or another MCP client. Your AI assistant can then search patents, read claims, follow prosecution history, and look up trademarks without anyone copying records out of registry websites.

Patents

USPTO in depth

Plus publications from other major offices

Trademarks

80+ offices

EUIPO and national registries worldwide

Format

Compact

Trimmed records, with long text paged

Example MCP call

Works with any MCP client

Prompt

Is US11267227B2 still in force, and has it been assigned?

Tool call

get_patent_legal_events { "patentNumber": "US11267227B2" }

Returned

Legal status, recorded events, and key dates

Follow-up

Claims, family, citations, prosecution history

ClaimCell for IP teams

  • Pull current patent and trademark records into an AI conversation instead of exporting them from each registry.
  • Read USPTO prosecution history, continuity data, and file-wrapper document lists from the official source.
  • Give everyone on the team the same tools for the questions they ask every week.

MCP product

Connect your AI tools to live IP data.

ClaimCell is an MCP (Model Context Protocol) server. Once it's added to your AI client, the model has a set of patent and trademark tools it can call on its own, so nobody has to switch between registry sites and paste results into the chat.

Ask in plain English about ownership, legal status, claims, citations, family members, Nice classes, or prosecution events. Answers are built from the office records, and each result carries the identifiers you need to check it at the source.

Supported workflows

Portfolio review

Check status, ownership, and filing history for the assets you care about.

Trademark searches

Search marks by text, owner, Nice class, office, and status before filing.

Drafting support

Bring claims, cited references, and family context into the draft you're working on.

Due diligence

Get a first read on status, ownership, and family coverage before the full review.

Consistent records

Records from different offices come back with predictable field names, so the model spends less effort working out what it's looking at.

Set up once

Add one endpoint and an API key to your MCP client. Everyone using that client gets the same tools.

How it works

From question to source record.

Request private beta access

01

Connect your MCP client

Add the ClaimCell endpoint and your API key to Claude, Cursor, Copilot, or another MCP client.

02

Ask your question

Ask in plain language. The model picks the right tool and fetches claims, classes, citations, events, or ownership details.

03

Check the record

Results include publication, application, and registration numbers, so you can confirm anything important with the office before relying on it.

Built for IP counsel, paralegals, and portfolio managers.

Use it for the questions that usually mean opening several registry tabs: what a patent claims, who owns it now, where an application stands in prosecution, and which marks are already registered in a class.

Use case

Competitor filings

Use case

Prosecution history

Use case

Continuity chains

Use case

Citation analysis

Technical detail

Small tools that each do one job.

Search, fetch a document, get a family, list prosecution events, look up a trademark. The model picks the tool that fits your question.

get_patent_family { "patentNumber": "WO2023198244A1" }

odp_get_transactions { "applicationNumber": "18123456" }

euipo_get_trademark { "applicationNumber": "018012345" }

Cost efficiency

Smaller responses, lower token bills.

Registry records are long. A full patent document or trademark file includes every field the office stores, and your AI client pays for each token it reads, useful or not.

ClaimCell trims each record to the fields that matter before sending it back. Long text such as claims and specifications arrives one page at a time, and the model asks for the next page only if it needs it.

  • Compact JSON records instead of raw XML or HTML pages.
  • Paged claims and descriptions, so a long specification doesn't fill the context window in one go.
  • Capped search results and citation lists, with totals so the model knows when there's more.

Same question, smaller payload

Without ClaimCell

The full office record, with every field, legacy tag, and bit of markup loaded into context.

With ClaimCell

get_patent_legal_events { "patentNumber": "US11267227B2" }

→ status, legal events, lifecycle dates

More of your context window goes to the question you actually asked.

Blog

Notes from building ClaimCell.

Product updates and how-to guides for IP and legal ops teams using ClaimCell with their AI tools.

Browse the blog

What we plan to publish

  • Walkthroughs for patent research, prosecution review, and trademark searches.
  • Short technical write-ups on search quality and source coverage.

Pricing

One plan, every tool.

Full access to patent and trademark data over MCP. No per-seat charges.

$500/month
  • Patent search, full text, families, citations, and legal events
  • USPTO prosecution history and file-wrapper documents
  • Trademark search across EUIPO and 80+ offices
  • Works with Claude, Cursor, Copilot, and any MCP client
  • Email support
Request private beta access

Cancel anytime. No long-term contracts.