ClaimHit

ClaimHit is an AI-powered patent claim-reading and infringement intelligence platform used by IP counsel, corporate intellectual property teams, patent prosecutors, litigation funders, patent owners, and universities and technology transfer offices (TTOs). The platform determines whether a patent's claims read on real, in-market products or on the mandatory provisions of technical standards, supporting use cases including enforcement, licensing and partnering, patent prosecution strategy, and litigation funding diligence.

Key features and functions include:

Multi-Source Candidate Discovery
Identifies candidate products and companies using several AI models in parallel, alongside live web and neural search, forward citations from companies whose own patents cite the patent being screened, and field-specific sources including component distributor catalogues, FDA drug and device records, software directories, online marketplaces, and first-party technical documentation such as datasheets, manuals, application notes, and FCC filings. Sources queried are matched to the patent's technology domain rather than searched uniformly.

Evidence Verification and Scoring
Each candidate is checked against citable evidence before being reported. Claim elements are individually scored as fully supported, partially supported, or unsupported ("NONE") based on the available record, without rounding up an unsupported limitation. Candidates confirmed directly on a manufacturer's own materials are distinguished from those identified through model agreement alone but not independently confirmed. Candidates that cannot be evidenced are excluded rather than shown with a caveat.

Vision-Based Evidence Reading
Uses a vision model to read block diagrams, schematics, and teardown photographs, and can incorporate user-uploaded documents such as teardown reports or internal specifications as additional evidence.

Patent Office Compatibility
Accepts patent numbers from major patent offices and fetches full claim text automatically from official databases.

Standard Essential Patent (SEP) Analysis
Maps patent claims against the mandatory provisions of technical standards, read from the underlying specification text of standards bodies including 3GPP, ETSI, IEEE, and ITU, rather than from summaries.

AI Hit Charts
Generates a claim chart mapping each claim element to a target product or standard provision, with supporting evidence and citations for each limitation, including limitations the platform could not evidence.

Litigation Support Screening
Provides pre-suit infringement screening intended to help litigators assess a candidate case before committing significant billable time, reporting both supported and unsupported claim limitations.

Patent Prosecution Support
Maps claims under prosecution, including published applications, against current in-market products or, for standard essential patents, against the specification text, to help prosecutors evaluate how a proposed claim amendment would affect claim scope before narrowing claim language.

Litigation Funding Diligence
Screens and ranks candidate infringement cases by the strength of available evidence, distinguishes confirmed matches from unconfirmed model-based matches, and supports batch screening of multiple candidate assets for comparison across a portfolio.

Expert Review
Provides human review of an AI-generated Hit Chart by a subject matter expert, who checks each claim element and evidence citation, with delivery within 36 hours. An optional consultation call with the review team is available.

Automated Monitoring
Allows a scan to be set to re-run automatically on a user-selected cadence (daily, weekly, monthly, or quarterly), with an email alert sent when a new match is identified. Each monitoring re-run uses one search credit.

MCP Server Integration
Available as a Model Context Protocol (MCP) server, allowing users to run searches, retrieve past scans, and check credit balances from within MCP-compatible AI assistants such as Claude and ChatGPT, using the same credits as the web application.

Team Collaboration
Supports shared team workspaces with a pooled credit system and shared search history.

Confidentiality
Search inputs are treated as confidential, are not sold, and are not used to train AI models, with each search isolated from other users' data.