Show HN: Omni – Local-first multimodal file search on macOS (hanxiao.io)

Omni, a new local-first multimodal file search application for macOS, has been announced, bringing a significant enhancement in how users can search for various file types, including text, code, PDFs, images, audio, and video, all within a unified vector space. This innovative approach enables users to execute queries across different languages (e.g., search in English and find files in German or Chinese) seamlessly while retaining complete control over their data, as every operation occurs on-device without requiring accounts or internet connectivity. The application leverages a native MLX-Swift implementation of the jina-embeddings-v5-omni model, running on Apple Silicon GPUs to provide fast, offline search capabilities. It features a local endpoint for semantic search, allowing agents like Hermes and OpenClaw to query files based on their meaning. Notably, Omni supports comprehensive file handling with functionalities like List and Gallery views, drag-and-drop features, and advanced filtering options. This release marks a significant move towards privacy-centric, robust file management tools in the AI/ML space, catering to users who prioritize security and efficiency in their digital workflows.