There are two main methods to install Perplexica locally:

1. Using Docker:

   • First, ensure you have Docker installed on your system.
   • Clone the Perplexica repository.
   • Modify the sample.config file with your API keys for OpenAI, Anthropic (AMA), and Cohere (Grok).
   • In the Perplexica directory, run the command docker-compose up -d to start the Docker container.

2. Without Docker:

   • Follow the installation instructions provided in the Perplexica documentation.
   • You can also easily deploy Perplexica using the Replit cloud service by clicking the "Deploy on Replit" button, creating an account, and providing the necessary API keys.

Once Perplexica is installed, you can access the web-based interface and start using the AI-powered search engine. You can customize the settings, such as selecting different language models, text embedding models, and API keys, to suit your specific needs.

Perplexica provides a powerful and flexible search experience, allowing you to ask questions, search the web, and receive concise and informative responses with relevant sources. Its architecture, which includes components like agents, chains, and embedding models, ensures accurate and up-to-date search results.

Perplexica's architecture comprises several key components that work together to provide an advanced AI-powered search engine experience:

1. User Interface: Perplexica offers a web-based platform that allows users to perform searches for text, images, and videos.

2. Agents and Chains: These components are responsible for predicting Perplexica's next action, understanding user queries, and deciding whether a web search is necessary.

3. Search Engine (SearXNG): Perplexica utilizes the SearXNG metasearch engine to score and retrieve relevant sources from the internet.

4. Large Language Models: Perplexica employs various large language models, including cloud-based and open-source options, to understand the content and generate responses.

5. Embedding Models: These models are used to enhance the accuracy of search results by re-ranking them using similarity search algorithms.

Together, these components work seamlessly to provide users with concise and detailed answers, while also ensuring the search results are up-to-date and accurate.

Perplexica is an open-source AI-powered search engine that offers a range of advanced features to enhance your search experience. Here are some of the key features that make Perplexica a powerful tool:

1. Local Language Model Integration: Perplexica allows you to utilize local large language models, enabling you to create a search engine tailored to your specific context or company needs.

2. Dual Search Modes: Perplexica offers two main modes of operation - the normal mode for standard web searches, and the Copilot mode, which is currently in development. The Copilot mode aims to enhance searches by generating diverse queries to find more relevant sources.

3. Specialized Focus Modes: Perplexica provides various focus modes to cater to different types of queries, including All Mode (searching the entire web), Writing Assistant Mode (for writing tasks without web searches), Academic Search Mode (for research papers and articles), YouTube Search Mode, and more.

4. Current Information Retrieval: Perplexica utilizes the SearchNG meta-search engine to retrieve and re-rank results, ensuring that you get the most up-to-date and relevant information, avoiding the common issue of outdated information found in systems relying on pre-crawled data.

5. Image and Video Search: In addition to text-based searches, Perplexica supports image and video search capabilities, expanding the scope of your queries.

6. Ongoing Development: Perplexica is continuously evolving, with plans to finalize the Copilot mode, add a discover and history-saving feature, and integrate various other enhancements in the future.

By leveraging these powerful features, Perplexica aims to provide a comprehensive and efficient search experience, empowering users to find the most relevant and up-to-date information with ease.

Perplex is an open-source, AI-powered search engine that aims to provide more accurate and up-to-date information compared to traditional search engines. It utilizes a combination of large language models, embedding models, and specialized search components to enhance the search experience.

Some key features of Perplex include:

• Support for local language models, allowing customization for specific use cases
• Two main modes: a normal mode for web searches and a Copilot mode for generating more relevant queries
• Specialized focus modes for different types of searches (e.g., writing, academic, YouTube, Reddit)
• Utilization of SearchNG, a meta-search engine, to retrieve and rank the most current information

Perplex can be installed either using Docker or a non-Docker approach, and it can also be deployed on cloud platforms like Replit. The open-source nature of Perplex allows for flexibility and customization, making it a potentially valuable tool for various applications that require accurate and up-to-date information retrieval.