Scira is a minimalistic, AI-powered search engine designed for users seeking direct answers with citations. It integrates multiple AI models and specialized search APIs to provide comprehensive information retrieval, aiming to be a more efficient and informative alternative to traditional search engines.

How It Works

Scira leverages the Vercel AI SDK for seamless integration and streaming of responses from various large language models, including xAI's Grok, Anthropic's Claude 3.5 Sonnet, and Meta's Llama 3.3. For grounding search results and performing web searches, it utilizes Tavily AI. The engine also incorporates specialized APIs for specific functionalities like weather (OpenWeather), code execution (E2B), mapping (Google Maps, Mapbox, TripAdvisor), and academic/X post searches (Exa.AI).

Quick Start & Requirements

• Local Development:
  • Docker Compose (Recommended): docker compose up (requires Docker and Docker Compose, .env file with API keys).
  • Node.js: pnpm install then pnpm dev (requires Node.js, pnpm, OpenAI, Anthropic, and Tavily API keys).
• Prerequisites: API keys for OpenAI, Anthropic, and Tavily are required for local development.
• Default Search Engine: Instructions are provided to set Scira as the default search engine in Chrome.
• Documentation: https://scira.ai/

Highlighted Details

• Supports multiple LLMs: xAI Grok, Anthropic Claude 3.5 Sonnet, Meta Llama 3.3 70B, Deepseek R1.
• Integrates specialized APIs for weather, code execution, maps, flight tracking, and academic/X post searches.
• Built with Next.js, Tailwind CSS, Vercel AI SDK, and Shadcn/UI.
• Offers YouTube search with timestamps and transcripts via Exa.AI.

Maintenance & Community

The project acknowledges Warp and NinjaTools. No specific community channels or roadmap details are provided in the README.

Licensing & Compatibility

• License: Apache 2.0 License.
• Compatibility: Permissive license suitable for commercial use and integration into closed-source projects.

Limitations & Caveats

The README focuses on setup and features, with no explicit mention of limitations, known bugs, or alpha/beta status. The reliance on multiple third-party APIs means service availability and costs are dependent on those providers.