docuchat
Local AI document chat checks citations against your legal files. Runs offline on Apple Silicon Macs with Ollama, free under the MIT license.

docuchat is a local AI document assistant for attorneys, especially solo and small firms working with confidential case files. It answers questions across contracts, pleadings, exhibits and transcripts, then checks cited quotes against the source text. Answers point to the file, page and exact passage; deposition citations include verified line numbers.
Documents stay on your computer. Inference, search, OCR and storage run locally, with no cloud calls in the question-and-answer process, no account requirement and no telemetry. It works offline after an initial model download. The desktop app runs on Apple Silicon Macs; a Windows desktop build is still forthcoming. Source deployment with Python and Ollama supports macOS, Windows and Linux. It's free and open source under the MIT license, using Ollama with qwen3:14b for answers and bge-m3 for embeddings.
Each matter keeps its own documents and conversations, and searches stay within that matter. Scanned PDFs go through local Tesseract OCR, while tables and exhibits become searchable content with citations. A transcript digest produces a Word table of verified verbatim quotes. For broader review, a grid compares cited answers across documents, and a contract checklist flags clauses as found or potentially missing.
Matter overviews show source-verified parties, timelines and deadlines. Attorneys confirm deadline dates before adding them to a calendar; the app doesn't calculate those dates itself. It locates and summarizes document content without giving legal advice or drawing legal conclusions. Optional imports from Gmail, Notion and Zoom contact those services to pull files in, while document processing stays local.