File Brain

File Brain is a local file search engine for the desktop. It indexes documents, images and other files on a computer and finds them by meaning rather than by filename, using vector embeddings for semantic matching and built in OCR to read text inside screenshots and scanned documents. Indexing and querying run entirely on the machine, so file contents are not sent anywhere, and it runs on Windows, macOS and Linux. It is built for people whose work spreads across thousands of files and who remember what a document said but not where it was saved. Search tolerates typos, matches concepts instead of exact keywords, and works across languages, so a query in one language can return documents written in another. The index covers a wide range of document, archive and binary formats, duplicate files are identified by content even when the names differ, and folders can be watched for automatic indexing. The free tier covers local search, typo tolerant and semantic text search, automatic indexing and command line use. The free version costs nothing and its source is published in a public repository. Paid tiers are perpetual licenses sold as a one time payment with one year of automatic updates and priority support, and they are taken as preorders rather than sold outright. Knowledge Engine has a list price of $79 and adds chat with files, finding similar files by text, and agent and MCP support. Media Suite is listed at $119 and adds visual understanding, including finding objects in images and scenes in video. Cloud Connect is listed at $169 and adds network drives, cloud storage and a remote web interface. Lower preorder prices are shown next to each list price. Buyers should note that the paid features are not released yet, so the preorder tiers describe planned rather than shipping functionality.