Perplexica: The AI Powerhouse for Your Search Needs
Overview
Perplexica is an open-source, AI-powered search engine designed to enhance the way people find information online. It's not just a regular search engine; it is capable of understanding questions and providing precise answers with cited sources, making it a standout tool for anyone looking to dive deep into the vast internet knowledge pool. Modeled after Perplexity AI, Perplexica employs cutting-edge machine learning techniques such as similarity searching and embeddings to refine results and bring clarity to your queries. Furthermore, it respects users' privacy by leveraging SearxNG, an open-source metasearch engine, to fetch the most current information.
Features
- Local LLMs: Perplexica supports local Large Language Models (LLMs) like Llama3 and Mixtral via Ollama, giving you flexibility in how you conduct your searches.
- Two Main Modes:
- Copilot Mode: (In development) This mode revolutionizes search by automatically generating varied queries to gather the most relevant information from the internet, analyzing multiple sources for a comprehensive response.
- Normal Mode: Simply processes your query and performs a web search.
- Focus Modes: Tailor your search with specific modes, including:
- All Mode: Scours the web comprehensively for the best results.
- Writing Assistant Mode: Assists in writing tasks without involving extensive web searches.
- Academic Search Mode: Searches for academic articles and papers, ideal for research purposes.
- YouTube Search Mode: Focuses on locating videos on YouTube relevant to your query.
- Wolfram Alpha Search Mode: Handles questions requiring computational answers or data analysis.
- Reddit Search Mode: Finds discussions and opinions on Reddit related to your topic.
- Current Information: Unlike tools that rely on periodically updated embeddings, Perplexica uses live results from SearxNG, ensuring you always have access to the latest information.
- API: Developers can integrate Perplexica's capabilities into their applications, leveraging its robust search functionalities.
Installation
Perplexica can be installed using two methods: via Docker (recommended) or through a manual setup.
Getting Started with Docker (Recommended)
- Ensure that Docker is installed and operational.
- Clone the Perplexica repository and adjust the necessary configuration files.
- Deploy the setup with a simple
docker compose up -dcommand. - Access Perplexica through your web browser at http://localhost:3000.
Non-Docker Installation
- Install SearXNG and configure it for JSON output.
- Prepare configuration files and install dependencies.
- Build and run the application by executing commands within the appropriate directories.
Additional Setup Details
Instructions for tackling Ollama connection issues and for using Perplexica as a search engine in your browser are provided in the project documentation. These guides help ensure a smooth integration and operation of Perplexica, even when faced with technical challenges.
Using Perplexica's API
Developers can tap into Perplexica's API to integrate its search capabilities within their own applications. Detailed documentation is available to help navigate its use.
Expose Perplexica to a Network
Perplexica can be set up to be accessible over a home network, offering the ability to use this powerful tool across different devices within the same network.
In summary, Perplexica offers a versatile, AI-driven search experience with features designed to meet diverse needs, from casual browsing to intensive academic research. The project's open-source nature invites collaboration and further development, ensuring that it continues to evolve and improve.