🛡️ Patents & Trademarks (IP) MCP Server

Give your AI agent instant access to worldwide patent intelligence. Search 130M+ patents, look up any publication by number, map a company's IP portfolio, and find an inventor's body of work — all through the Model Context Protocol, straight from Claude, Cursor, Windsurf, or any MCP client.

Built by NexGenData. Connect once, then your agent can run patent landscaping, freedom-to-operate scoping, prior-art discovery, and competitor IP monitoring on demand.

---

⚡ What you get

Tool	What it does	Data source
search_patents	Keyword / phrase / inventor / assignee patent search	Google Patents
get_patent_details	Full metadata for a specific publication number	Google Patents
find_patents_by_assignee	A company's patent portfolio	Google Patents
search_inventor	An inventor's full body of work	Google Patents

Every result includes patent ID, title, assignee, inventor, priority/filing/grant/publication dates, a relevance snippet, and a link to the source.

---

🎯 Use cases

• Patent landscaping — map the prior art in a technology area before R&D investment.
• Competitive IP intelligence — pull a rival's entire patent portfolio in one call.
• Freedom-to-operate scoping — surface relevant patents before launching a product.
• Prior-art / novelty checks — find existing filings that overlap an invention.
• Inventor & expert discovery — identify the people behind key innovations for hiring or licensing.

---

🚀 How to connect

Add this to your MCP client config (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "patents-ip": {
      "url": "https://nexgendata--patents-trademarks-ip-mcp-server.apify.actor/mcp",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

Get a free Apify token at apify.com (sign-up includes free monthly platform credit).

---

📊 Sample inputs

search_patents(query="solid-state battery electrolyte", max_results=10)
find_patents_by_assignee(assignee="NVIDIA", max_results=15)
get_patent_details(patent_id="US10000000B2")

📦 Sample output (search_patents)

{
  "query": "solid-state battery electrolyte",
  "total_results": 48213,
  "patents": [
    {
      "patent_id": "US11069918B2",
      "title": "Solid-state electrolyte for lithium battery",
      "assignee": "Toyota Motor Corp",
      "inventor": "…",
      "publication_date": "2021-07-20",
      "snippet": "A solid-state electrolyte comprising…"
    }
  ],
  "source": "Google Patents"
}

---

🧠 How it works

This is a standby MCP server, not a one-shot scraper. It boots once, holds an HTTP/MCP session open, and answers tool calls live. Each tool queries Google Patents' public search backend and normalizes the response into clean, agent-friendly JSON. No API key for the data source is required.

---

🔗 Related NexGenData Actors

Part of the NexGenData IP & Patent suite — pair this Actor with:

• Trademark Search USPTO Database Lookup
• USPTO Patent Search Claims & Prior Art
• USPTO Patent Grants Tracker IP & Competitive Intelligence
• USPTO Patent Grants Tracker Recent US Patent Filings
• EUIPO Trademark Search TMview & eSearch
• CNIPA China Patent Search IP & Innovation Tracker
• WIPO PATENTSCOPE Patent Search
• HK Trademark Search

💵 Pricing

Pay-per-event: you are charged a small fee per successful tool call, plus a one-time actor-start. No subscription. See the live pricing tab on the actor page for current rates. Build your own automations and run them through Apify.

---

⚖️ Legal

This server queries publicly available patent metadata. Patent documents and bibliographic data are public records. Use responsibly and in accordance with the data source's terms. Nothing here is legal advice — consult a qualified patent attorney for FTO or validity opinions.

---

❓ FAQ

Q: Does it cover non-US patents? Yes — Google Patents indexes US, EP, WO, JP, CN, KR and many other jurisdictions.

Q: Do I need a patent-office API key? No. The server handles data access for you; you only need your Apify token.

Q: Can it return full claims / full text? It returns rich bibliographic metadata and an abstract snippet plus a direct link to the full document. For full-text claims, follow the google_patents_url.

Q: Is this real-time? Results reflect the live Google Patents index at query time.

Q: How many results per call? Up to 30 for searches; tune with max_results.

Q: What about trademarks? Patent tools are live today; trademark coverage rolls out via sibling NexGenData actors (see Related Actors).

Q: Does it work with Claude / Cursor / Windsurf? Yes — any MCP-compatible client over streamable-HTTP.

---

🛠️ Troubleshooting

• No results → broaden the query or check the assignee/inventor spelling.
• Empty get_patent_details → verify the publication number format (e.g. US9876543B2).
• Rate limiting → the upstream index may throttle bursts; retry after a moment.
• Auth error → confirm your Apify token is set in the Authorization header.
• Connection drops when idle → the server closes idle sessions to save compute; just reconnect.

---

🏢 About NexGenData

NexGenData builds a fleet of production MCP servers and data actors for AI agents — finance, legal, regulatory, IP, government, aviation, and more. Explore the full catalog on Apify.