Lepton Search – open source conversational AI search engine project

What is Lepton Search?

Lepton Search is an open source conversational AI search engine of the Lepton AI application construction platform founded by former Alibaba technology vice president and AI scientist Jia Yangqing. The project is based on calling the cloud artificial intelligence model on the Lepton platform and the API key of Bing search , a conversational search engine similar to Perplexity AI was built with less than 500 lines of Python code .

Features of Letpton Search

• Free and open source : The code of Lepton Search is open source, which means that developers can freely use, modify and distribute the code, promoting the sharing and innovation of technology.
• Conversational search : Lepton Search allows users to interact with search engines through natural language, providing a more intuitive and friendly search experience.
• Built-in large language model (LLM) support : Lepton Search integrates large language models, which enables it to understand and generate natural language to provide more accurate search results.
• Search engine integration : Lepton Search has a built-in search engine function that can quickly retrieve and provide relevant information by calling an external search engine API (Bing Search API).
• Customized UI interface : Developers can customize the user interface of Lepton Search according to their own needs to adapt to different application scenarios and user preferences.
• Search result sharing and caching : Lepton Search supports sharing search results and can cache search results to improve user experience and search efficiency.
• Cloud-native platform : Lepton Search runs on Lepton AI’s cloud platform, which has scalability, flexibility and security, and is suitable for handling large-scale AI tasks.

Lepton Search’s official website entrance

• Official website address: https://search.lepton.run/
• GitHub code base: https://github.com/leptonai/search_with_lepton

How to use Lepton Search

Ordinary users can experience the Lepton Search online web version, and developers can also deploy it themselves. The specific steps are as follows:

1. Clone the Lepton Search project:git clone https://github.com/leptonai/search_with_lepton.git
2. Run pip install -U leptonai && lep loginthe command to automatically set up the LLM and KV functions
3. Set up Bing key:export BING_SEARCH_V7_SUBSCRIPTION_KEY=YOUR_BING_SUBSCRIPTION_KEY
4. Build the front-end network:cd web && npm install && npm run build
5. Run the background server:BACKEND=BING python search_with_lepton.py