LexiFinder is an AI-powered tool that automatically generates analytic indexes for books, theses, and manuscripts. Instead of marking every term by hand, you simply provide your document and a list of keywords — LexiFinder does the rest. LexiFinder reads your document, extracts all nouns, and uses semantic analysis to find every term related to your keywords, even when the exact word does not appear. The result is a structured, hierarchical index with page or paragraph references, ready to be included in your publication. Supported input formats are PDF, DOCX, and ODT. The index can be exported as plain text, HTML, JSON, or CSV. LexiFinder produces a two-level index by default: your keywords become the main entries, and semantically related terms become sub-entries. A hybrid mode creates a three-level hierarchy with automatic sub-grouping. Fully automatic indexing modes are also available and require no keywords at all. For DOCX and ODT files, LexiFinder can mark the document directly with index entry fields, so that Word and LibreOffice can generate their own native index. A batch mode lets you process entire folders of documents in a single run. Reusable configuration files make it easy to apply the same settings across a multi-chapter project. The app supports English, Italian, French, German, and Spanish, and adapts to your system's dark or light theme. LexiFinder also works as a command-line tool for scripting and automation. LexiFinder is designed for authors, academics, editors, and publishers who need to produce a back-of-book index without spending hours on manual markup. LexiFinder is free and open source software, released under the GNU General Public License v3 (GPLv3). The source code is available at https://github.com/andreaciarrocchi/lexifinder. For more information, visit https://andreaciarrocchi.altervista.org

Automatically generates analytic indexes from PDF, DOCX, and ODT documents Semantic analysis finds related terms even when the exact keyword is absent Two-level hierarchical index: main keywords and semantically related sub-entries Three-level hybrid mode with automatic sub-grouping Fully automatic indexing modes require no keywords Marks DOCX and ODT files with index entry fields for Word and LibreOffice Batch mode processes entire folders in a single run Exports to plain text, HTML, JSON, and CSV Supports English, Italian, French, German, and Spanish Reusable JSON configuration files for multi-chapter projects Built-in model manager to download and delete language models Command-line interface for scripting and automation Dark and light theme support Free and open source (GPLv3)