DOCSEARCHbyAlgolia
Submit Docs

TurboSeek Documentation

repository·main·Indexed 23 days ago

https://github.com/nutlope/turboseek

An open source AI-powered search engine using Retrieval-Augmented Generation (RAG) to provide answers based on real-time web search results. It integrates Together.ai for LLM inference (gpt-oss and Llama 3.1 8B) and utilizes Bing or Serper APIs for web searching and content scraping.

Tokens
225
Snippets
1
Records
2
Agent score
32%

What's inside TurboSeek

  1. Overview of TurboSeek

    main
    TurboSeek is an open source AI search engine powered by Together.ai. It functions by taking a user's question, performing a web search via Bing/Serper API, scraping the content from the top results to use as context, and then using LLMs (gpt-oss and Llama 3.1 8B) to generate a streamed answer and suggested follow-up questions.

  2. Clone and run TurboSeek locally

    main

    To run TurboSeek on your local machine, follow these steps:

    1. Fork or clone the repository.
    2. Obtain API keys for the following services:
      • Together AI: For LLM inference.
      • SERP API (serper.dev) or Azure Bing Search API: For web search capabilities.
      • Helicone: For observability.
    3. Create a .env file in the root directory using .example.env as a template and populate it with your API keys.
    4. Install dependencies and start the development server using npm.
    npm install
npm run dev