XiaoyaoSearch is an AI-powered local file search tool designed for knowledge workers, content creators, and developers, offering multi-modal input and advanced AI models for efficient and intuitive file retrieval.
Source: README View on GitHub →XiaoyaoSearch is gaining attention due to its innovative multi-modal input capabilities, integration of advanced AI models like BGE-M3 and CN-CLIP, and its focus on privacy and performance. It fills the gap in local file search tools by providing a user-friendly, AI-enhanced solution.
Source: READMESupports voice, text, and image inputs to convert user queries into semantic searches for local file retrieval.
Source: READMEIntegrates AI models for semantic understanding and advanced search capabilities, including support for video, audio, and document content and file name search.
Source: READMEUtilizes a hybrid search architecture with Faiss and Whoosh, offering both performance and privacy options, with local data not automatically uploaded.
Source: READMEThe architecture is a hybrid of a client-server model with Electron for the frontend and FastAPI + Uvicorn for the backend. It uses Faiss for vector search and Whoosh for full-text search, with SQLite for database storage.
Source: Code tree + READMECenter: project; inner ring: core feature modules; outer ring: key dependencies. Auto-generated from core_features and tech_stack.key_deps.
FaissWhooshSQLiteXiaoyaoSearch is suitable for knowledge workers, content creators, and developers who need an efficient and intuitive way to search local files, especially those who require multi-modal input and AI-enhanced search capabilities.
Source: READMEv2.0.0 (2026-04-24): Major UI upgrade with Notion warm and bright design style and complete design system specification.
Source: GitHub ReleasesXiaoyaoSearch is a promising project for teams or individuals seeking an advanced, AI-powered local file search solution. Its unique multi-modal input and AI integration make it stand out, particularly for users who require efficient and intuitive file retrieval with a focus on privacy and performance.