AI & LLM

Perplexica: AI-enhanced Next-generation Search Engine

0

In modern society, information search is an essential part of daily life. However, traditional search engines still have limitations. To address this issue, Perplexica has emerged, leveraging AI technology to provide more accurate and reliable search results. In this article, we will explore the main features of Perplexica, its installation method, and usage scenarios.

1. Main Features of Perplexica

Unlike traditional search engines, Perplexica focuses on understanding user queries through AI and providing deeper answers. It uses SearxNG to offer the latest information while protecting user privacy.

Using Local LLMs

You can use local LLMs such as Llama3 and Mixtral.

Two Main Modes

It is divided into Copilot mode and Normal mode.

  • Copilot Mode: Enhances searches to find more relevant internet sources.
  • Normal Mode: Performs general web searches.

Focus Modes for Specific Queries

There are six focus modes.

  • All Mode: Searches the entire web.
  • Writing Assistant Mode: Helps with writing without web searches.
  • Academic Search Mode: Finds academic papers and articles.
  • YouTube Search Mode: Searches YouTube videos.
  • Wolfram Alpha Search Mode: Answers questions requiring calculations or data analysis.
  • Reddit Search Mode: Searches discussions and opinions on Reddit.

2. Installation Method

Installing Perplexica via Docker is highly recommended. Using Docker simplifies the management of environment variables and dependencies.

Installation Using Docker

1. Install and Run

Install and run Docker.

2. Clone Repository

Clone the Perplexica repository:

git clone https://github.com/ItzCrazyKns/Perplexica.git
3. Fill Required Fields

Move to the project directory, rename `sample.config.toml` to `config.toml`, and fill in the required fields.

4. Execute

Run Docker Compose:

docker compose up -d
5. Open Browser

Enter `http://localhost:3000` in your browser to access Perplexica.

Installing Without Docker

Perplexica can also be installed without Docker.

  • Clone the repository and rename `sample.config.toml` to `config.toml`.
  • Rename `.env.example` to `.env` in the `ui` folder and fill in the fields.
  • Run `npm i` in the `ui` folder and root directory to install dependencies.
  • Run `npm run build` to build.
  • Run `npm run start` to start the frontend and backend.

3. Usage Scenarios of Perplexica

If you want to use Perplexica as an alternative to traditional search engines, you can quickly access it through browser settings.

  • Open browser settings.
  • Go to the ‘Search Engine’ section.
  • Add a new site search: `http://localhost:3000/?q=%s`
  • Click the add button to use Perplexica directly from the browser search bar.

Perplexica is continuously evolving and will improve further through your contributions and feedback. If interested, star the project on GitHub and join the Discord server to interact with other users.

References: Perplexica, “An AI-powered search engine”

Leave a Reply