# Who uses Antecedint People # Used, and corrected, by the people who search it. Prosecutors, litigators, analysts and legal AI teams can all search the same public record here, free. Report a wrong or missing record, and once a person accepts it, the fix is in the next build for everyone. Who uses it ## Four kinds of work, one index. ### Patent prosecutors Prior-art search before you draft, and claim-by-claim coverage before you file. Every hit carries the source passage, so an office-action response starts from the record instead of a blank page. ### Litigators & IPR Invalidity search that ranks the closest art and shows which claim elements each reference reads on. Build the chart from scored, cited references, not keyword luck. ### Legal AI teams Search under your own product. Call the MCP server or the API, free to start, and get cited results back. Ship the product instead of building a citation warehouse first. ### IP operations & analysts Landscape a field, resolve assignees, and see where art clusters. The same index, read for portfolio work instead of a single claim. MCP Feedback open ## Report what is wrong. We fix it. Antecedint is early, and the people who use it shape it. When you or your assistant find a wrong or missing record, report it with submit_data_feedback. A person reviews every report, and accepted ones go into the next build of the data garden. ### Shape the roadmap Tell us which corpus and which workflow you need next; what people ask for moves up the queue. ### Credits for every accepted fix An accepted report earns 10, 50 or 200 credits, by how much of the data it fixes. Reporting is free, up to 100 a day. ### Your questions stay yours We do not log or store query text, and never train on it. The search box on the site is a demo; the main way to use Antecedint is from the assistant you already use. No testimonials here. The benchmark is published with its harness, so you can rerun it yourself. ## Search it yourself. Free from your AI assistant over MCP, with no daily limit. Report what is wrong as you go.