ParancU

ParancU helps you find evidence inside your own documents. Import a PDF, paste text, scan a page, or use an image to create a local corpus. Then ask questions in natural language and retrieve the passages most relevant to what you are looking for. Rather than hiding retrieval behind a generated answer, ParancU keeps the evidence visible. Results expose candidate passages, guiding questions, summaries, and source text so you can inspect why a passage was retrieved. ParancU combines semantic and lexical retrieval, with an adjustable balance between the two. Multilingual embeddings allow questions and source documents to use different languages. Designed around a local-first architecture, document retrieval and embeddings run on the device without requiring a remote embedding service. KEY FEATURES • Import PDF documents or paste text • Scan documents or import images • Ask questions in natural language • Multilingual document retrieval • Semantic + lexical evidence ranking • Adjustable Semantic-Lexical Ratio • Inspect multiple candidate passages • View guiding questions, summaries, and source evidence • Local embedding inference and persistent local corpora ParancU evolved from the EcoSearch research prototype, exploring evidence-first retrieval: finding passages that contain useful evidence rather than relying on semantic similarity alone.