An agentic research platform that plans, retrieves, and cites information from the web and various APIs.

Scira is an open-source, self-hosted web app and server that provides AI-generated responses to queries. It functions as an agentic research platform that plans, retrieves, and cites information from various live sources. The software is deployed as a web application and can be configured as a default browser search engine to replace traditional search providers.
Users interact with the system by typing questions, uploading PDF files, or pasting URLs. The agent breaks complex queries into sub-tasks, searches live data, and provides grounded answers with inline citations. It supports a wide range of LLM providers, including OpenAI, Anthropic, Google, DeepSeek, and xAI, allowing users to select the model that best fits their specific research needs.
The architecture is built with Next.js, Tailwind CSS, and the Vercel AI SDK. It utilizes a variety of third-party APIs for specialized data retrieval, such as CoinGecko for cryptocurrency, TMDB for entertainment metadata, and OpenWeatherMap for environmental data. The system uses Drizzle ORM for database management, Upstash for serverless Redis rate limiting, and Cloudflare R2 for object storage of artifacts. It is designed for researchers, developers, and power users who require a customizable search interface with the ability to bring their own AI models and connect custom tools.
Scira is licensed under AGPL-3.0, positioning it as an extensible and open tool for those who prefer self-hosting their research infrastructure.
A privacy-focused AI answering engine that combines web search with local or cloud based language models.
An AI search engine that generates grounded answers with interactive UI components instead of plain markdown text.
Join our newsletter to get shiny new open source software delivered to your inbox. Unsubscribe anytime.