FindIP

Findip provides patent search for AI workflows through a browser-based playground and REST API for embedding search in other services. Its database covers published patents from the United States, China, Japan, Korea, and Europe. Full-text search covers publications from 2001 onward, and China from 2010; unpublished patents are excluded. Semantic search matches by meaning instead of exact keywords. Queries can be written in English, Korean, Japanese, Chinese, or German across the five jurisdictions. Other search options find visually similar patent drawings or similar DNA, RNA, and protein sequences; image and sequence matches can be reranked by topic. Findip says results correspond to existing publications that can be checked by publication number. Its MCP connection supports Claude Desktop and Web and ChatGPT during beta. The Free plan includes 30 search credits and all APIs. Text search uses one credit; image search, sequence search, and trend analysis use two. Basic starts at $45 a month, excluding tax, with 300 monthly credits. Free, Basic, and Pro list a rate limit of 60 requests per minute.

Findip offers several ways to search published patent records, with a free tier and API access. Its coverage excludes unpublished patents, and credit use varies by search type.

Findip may suit users searching published patents across its five covered jurisdictions, including through semantic, drawing, or sequence searches. Its API and MCP connection may suit teams integrating patent search into other services or workflows.

Overview

FindIP is a patent-search service with a browser playground and a REST API for adding search to AI workflows. It suits researchers and product teams exploring patents by concept, drawing, or biological sequence. Its range of search modes is useful, but the published-record boundary and credit-based usage make it a poor fit for work that requires unpublished filings or unrestricted searching.

Key features

Search by meaning, image, and sequence

Semantic search finds patent text by meaning rather than exact wording, which can help when a query uses different terminology from a publication. Queries in English, Korean, Japanese, Chinese, or German can reach across all five covered jurisdictions. FindIP says each result is an existing publication in its database and can be checked by publication number, a practical way to verify a result's identity.

Drawing search finds patents with visually similar figures; pairing an image with text can rank matches around a topic. Sequence search finds similar DNA, RNA, or protein sequences and can also rerank results by topic. These modes make FindIP more relevant to visual and life-science searches than a text-only workflow. Claim analysis, patent clustering, and family analysis are also supported.

Coverage and integrations

Coverage includes published patents from the US, China, Japan, Korea, and Europe. Full-text search reaches publications from 2001 onward, except in China, where it begins in 2010. Because the data comes from published office records, unpublished patents are outside the service's reach.

The browser-based playground gives users a direct way to work with search, while the REST API is intended for embedding it in other services. API keys are issued on request, with endpoint documentation provided when a key is issued. FindIP's MCP connection officially supports Claude Desktop and Web, and ChatGPT during beta, through the shared endpoint mcp.findip.ai.

Pricing

The Free plan costs nothing and includes 30 search credits, all APIs, five jurisdictions, and re-ranking. Community support is the trade-off. It is a reasonable way to explore the service, but 30 credits can run down quickly: text search costs one credit, while image search, sequence search, and trend analysis cost two each. Details, drawings, document fetches, and MCP resources cost no credits.

Basic costs $45 a month, billed Excl. tax, for one user with 300 search credits/month, all APIs, five jurisdictions, re-ranking, and email support. It adds a meaningful allowance and direct support over Free without team administration. Team Basic costs the same $45 a month, billed Excl. tax, and provides 300 search credits/member/month, a team dashboard, unified billing, self-service seat changes, and five jurisdictions. Teams that need shared administration should prefer it; solo users get more from Basic's included all-API access and re-ranking.

Pro costs $135 a month, billed Excl. tax, for a single user with 1,000 search credits/month, all APIs, five jurisdictions, re-ranking, and priority support. Team Pro costs $135 a month, billed Excl. tax, and instead offers 1,000 credits/member/month, a team dashboard, unified billing, self-service seat changes, and priority support. The higher allowance suits heavier search workloads; the team version is the clearer fit when multiple members need their own monthly credits. Both remain subject to a listed rate limit of 60 requests per minute.

Enterprise has custom pricing, search credits, contract and billing, support, and features or infrastructure by agreement; its rate limit is also custom. It is the option for organizations that need negotiated terms rather than a fixed plan. No paid plan removes the published-patent coverage limit.

Platforms

FindIP supports Android, iOS, web, and API access. The rate limit is 60 requests per minute on Free, Basic, and Pro; Enterprise uses a custom limit.

Who it's for

FindIP is best for people researching published patent records across several jurisdictions who can benefit from semantic, drawing, or sequence search, and for teams embedding patent search in AI workflows. It is less suitable when unpublished filings are essential or when a workflow needs more than the five covered jurisdictions.

Pros and cons

• Pros: Semantic, drawing, and biological-sequence search provide distinct routes into patent records beyond keyword queries.
• Pros: Five-jurisdiction coverage and queries in five languages support cross-border exploration.
• Pros: Zero-credit access to details, drawings, document fetches, and MCP resources helps conserve search credits.
• Cons: Unpublished patents are excluded, limiting its value for work that depends on filings before publication.
• Cons: Image, sequence, and trend searches each consume two credits, so specialized exploration uses the allowance faster than text search.
• Cons: API keys require a request, adding a step before developers can integrate the service.

Alternatives

For a broader comparison, browse AI Patent Search Software.

• Relecura Novelty is a paid option on web, iOS, and Android; consider it if those platforms suit your workflow.
• Patently offers a free web plan for individual use with asset search filters, patent-family browsing, and chat with Onardo; choose it if those free features are a closer match.
• PatSnap has a free plan and trial, with web, API, and self-hosted platforms; it may suit readers comparing those access options.
• GoVeda offers a free-forever OpenPatent plan with worldwide patent coverage and unlimited AI semantic searches; consider it if unlimited semantic search is the priority.
• The Lens offers free anonymous browsing and a free academic workspace; it may fit readers seeking those access routes.
• AcclaimIP is a paid web and API option with a free trial and per-seat annual licensing; consider it if that licensing structure suits your organization.
• Orbit Intelligence is a paid web option with Advanced and Essential plans; consider it if you are comparing those plan choices.
• XLSCOUT is a paid web and API option with a free trial; it is another service to compare.

Verdict

Choose FindIP if your work benefits from semantic, visual, or sequence-based searches across published patents and you value API or MCP access. Its strongest case is the breadth of search modes; its clearest reason to look elsewhere is that unpublished records are excluded and specialized searches use credits quickly.