Patent Search MCP — AI Agent Patent Intelligence
Pricing
Pay per usage
Patent Search MCP — AI Agent Patent Intelligence
Search patents across USPTO, EPO, and Google Patents for AI agents. One tool call to find patents, citations, and company landscapes.
Pricing
Pay per usage
Rating
0.0
(0)
Developer
AutomateLab
Actor stats
0
Bookmarked
2
Total users
1
Monthly active users
6 hours ago
Last modified
Categories
Share
Patent Search MCP Server
Search patents across USPTO, EPO, and Google Patents for AI agents.
What It Does
Give AI agents the ability to search patent databases, analyze patent landscapes, and trace citation chains — with one tool call.
- USPTO patents — full US patent database search
- EPO patents — European Patent Office search
- Google Patents — aggregated patent data with citations
- Company landscapes — full patent portfolios by assignee
- Citation tracking — forward and backward citations
Quick Start
Add to your AI agent:
{"mcpServers": {"patent-search-mcp": {"url": "https://patent-search-mcp.apify.actor/mcp"}}}
Or with authentication:
{"mcpServers": {"patent-search-mcp": {"url": "https://patent-search-mcp.apify.actor/mcp?token=YOUR_APIFY_TOKEN"}}}
Tools
| Tool | Price | Description |
|---|---|---|
search_patents | $0.05 | Search by keyword, CPC code, or inventor |
get_patent_details | $0.03 | Full metadata, claims, assignee |
find_patent_citations | $0.05 | Forward/backward citation chain |
patent_landscape_by_company | $0.10 | Company patent portfolio + filing trends |
Example Calls
Search Patents
search_patents(query="neural network transformer", max_results=10)
Returns:
{"patent_number": "US10712345B2","title": "Attention mechanism for neural networks","inventors": ["Vaswani", "Shazeer"],"filing_date": "2017-06-12","issue_date": "2020-08-25","assignee": "Google LLC","source": "USPTO","url": "https://patents.google.com/patent/US10712345B2"}
Get Patent Details
get_patent_details(patent_number="US10123456", source="all")
Returns:
{"patent_number": "US10123456","title": "Medical device with sensor array","abstract": "A medical device comprising...","assignee": "Medtronic PLC","source": "Google Patents","url": "https://patents.google.com/patent/US10123456","details_available": true}
Find Citations
find_patent_citations(patent_number="US10712345B2", citation_type="forward")
Returns:
{"patent_number": "US10712345B2","forward_citations": [{"patent_number": "US11012345", "source": "forward_citation"},{"patent_number": "EP3456789", "source": "forward_citation"}],"total": 156}
Company Landscape
patent_landscape_by_company(company_name="Apple Inc", max_results=20)
Returns:
{"company_name": "Apple Inc","total_patents": 247,"filing_trend": {"2020": 45, "2021": 52, "2022": 61, "2023": 89},"top_patents": [...],"technology_areas": ["semiconductor", "display", "wireless"],"sources_searched": ["USPTO", "Google Patents"]}
How It Works
Phase 1: Multi-Source Search
- Queries USPTO Patent Public Search API
- Queries Google Patents
- Queries EPO Open Patent Services (OPS)
- All queries run in parallel for speed
Phase 2: Deduplication
- Removes duplicate patents by number
- Preserves first-seen metadata
Phase 3: Aggregation
- Returns structured JSON with source attribution
- Includes direct URLs to patent records
Data Sources
| Source | Coverage | Type |
|---|---|---|
| USPTO | 12M+ US patents | Full text search |
| EPO | 100M+ worldwide | Patent families |
| Google Patents | Aggregated | Citations, assignments |
Use Cases
Tech Transfer Evaluation
"Find patents from MIT researchers in the AI space"
→ AI calls search_patents with MIT inventor query → Returns MIT patent portfolio
Due Diligence
"What's Apple's patent strategy in electric vehicles?"
→ AI calls patent_landscape_by_company → Returns filing trends, top patents
Citation Analysis
"Who has cited the original transformer patent?"
→ AI calls find_patent_citations with citation_type="forward" → Returns citing patents
Freedom to Operate
"Are there blocking patents on our surgical robot design?"
→ AI calls search_patents with design keywords → Returns relevant patents
Pricing
| Tool | Price |
|---|---|
search_patents | $0.05/call |
get_patent_details | $0.03/call |
find_patent_citations | $0.05/call |
patent_landscape_by_company | $0.10/call |
No subscription required. Pay per use via Apify PPE.
Cross-Sell Triggers
This MCP complements the research intelligence cluster:
university-research-mcp→patent_landscapetriggers →patent-search-mcpacademic-research-mcp→find_citationson patents →find_patent_citationshealthcare-compliance-mcp→ FDA device approvals → patent history check
Connect to AI Agents
Claude Desktop
{"mcpServers": {"patent-search-mcp": {"url": "https://patent-search-mcp.apify.actor/mcp"}}}
Cursor / Windsurf
Add the same JSON to your AI client config.
cURL Example
curl -X POST "https://patent-search-mcp.apify.actor/mcp" \-H "Content-Type: application/json" \-d '{"tool": "search_patents", "params": {"query": "neural network", "max_results": 5}}'
Output Schema
All tools return JSON. Each result includes:
patent_number— unique identifiertitle— patent titlesource— which databaseurl— direct link to record
API Status
- Health: Running
- Uptime: 99.9%
- Rate Limits: Respect upstream API limits
- Support: Open issue on GitHub