ReactiveSearch

ReactiveSearch is an open-source stack that helps developers build fast, AI-powered search experiences for websites and apps. It gives you the tools to create search UIs, connect to search engines like Elasticsearch and OpenSearch, and layer in AI models such as OpenAI for natural language queries and answers. The stack includes low-code UI kits for React, Vue, Flutter, and plain JavaScript, so you can ship a polished search interface without building everything from scratch.

What makes ReactiveSearch stand out is its composable search pipeline — you can write JavaScript pipelines that transform search requests and responses, mix backends, and inject AI logic. It also includes built-in analytics that track searches, filters, and results, giving you recommendations to improve relevance and fill content gaps. The platform supports self-hosting with Docker, comes with granular access controls using Basic Auth and JWT, and has an active open-source community on GitHub with over 15,000 stars.

ReactiveSearch is best suited for developers and product teams who need to add or upgrade site search — whether for an e-commerce catalog, a documentation site, or a content-heavy web app. It's also useful for teams that want to experiment with AI search features like Q&A, using their own data, without getting locked into a proprietary service. If you need search that's fast, customizable, and you can run yourself, this is a solid option.