Open Source TypeScript Search Engines

Scira is an open source AI-powered search and research assistant designed to provide fast, conversational answers grounded in web and knowledge sources. The project combines a modern web interface with retrieval-augmented generation techniques to deliver responses that are both natural language friendly and evidence oriented. It is built for developers who want to deploy their own Perplexity-style or AI search experience without relying on proprietary hosted services. Scira emphasizes speed, clean UI design, and extensibility so teams can customize data sources, models, and ranking logic. The architecture typically supports real-time querying, streaming responses, and modular backend components. Overall, Scira targets builders who want a self-hosted AI search experience focused on transparency and customization.

Web Page Application: https://shemeshg.github.io/desktop-search/ Github: https://github.com/shemeshg/desktop-search-code * Dropbox sync is performed manually (from admin screen) and not on interval.

memfree is an open source hybrid AI search engine and page generation platform designed to help users retrieve information from both personal knowledge bases and the public web through a unified interface. The project combines retrieval-augmented search with AI summarization to deliver concise answers instead of forcing users to manually sift through multiple sources. It supports multiple AI models and search providers, enabling flexible configuration depending on cost, performance, or privacy requirements. Beyond search, memfree includes an AI Page Generator that can transform text or images into production-ready frontend code using modern web stacks. The platform emphasizes productivity by automatically organizing information and reducing the need for manual bookmarking or note management. Overall, memfree positions itself as an all-in-one knowledge assistant and rapid UI generation tool for developers and power users.

Reverse Image Search Anywhere is a free, open-source (MIT) browser extension for Chrome, Firefox, Edge and Opera. Most reverse-image tools work by handing a search engine the image's URL. That breaks the moment a site hides it. Instagram and Pinterest do, and so does anything drawn to a canvas or served as a blob: or data: URL. They also make you check one engine at a time, which is the slow half of the job. This extension fixes both halves. Right-click any image and search Google Lens, Yandex, Bing Visual and TinEye, one engine or all of them at once. It also handles clipboard paste, region capture, and a face-search-first flow for finding people. Privacy: plain searches never touch our servers. Uploads happen only when you ask for them, get a link that expires after 15 minutes, and the file is deleted automatically about a day after it arrives. The source is published here so you can check that rather than take it on trust.

Reverse image search the way it should work: one right-click, every engine, even where the usual trick fails. Most "search by image" tools break on Instagram, Pinterest, and images drawn to a canvas or served as blob:/data: URLs, and make you check each engine one at a time. This extension fixes both. Right-click any image to search Google Lens, Yandex, Bing Visual and TinEye at once, or pick a single engine. You can also paste an image (Ctrl/Cmd+V), drag to capture a region of the page, or crop and run a dedicated "Search this face" flow. Private by default: ordinary searches go straight from your browser to the engine and never touch our servers. Where a site hides the image URL, the extension uploads that one image to mint a temporary link; it auto-deletes within about an hour and is never sold or used for training. No account, no tracking, no telemetry. MIT licensed. Manifest V3, built with WXT and TypeScript. Chrome and Firefox.