OpenPerPlex is an open-source AI search engine that leverages cutting-edge technologies to provide search capabilities over the web.
https://github.com/YassKhazzan/openperplex_front.git
- Semantic chunking using Cohere and semantic-chunkers library (
https://github.com/aurelio-labs/semantic-chunkers/blob/main/semantic_chunkers/chunkers/statistical.py
) - Reranking results with JINA API
- Google search integration via serper.dev
- Groq as the inference engine
- Llama 3 70B MODEL
- Python 3.11+
- pip
- Clone the repository: git clone https://github.com/YassKhazzan/openperplex_backend_os.git
- Install the required packages:
pip install -r requirements.txt
- Set up your environment variables:
- Copy the
.env_example
file to.env
- Fill in your API keys in the
.env
file
To start the OpenPerPlex server: uvicorn main:app --port 8000
The server will be available at http://localhost:8000
Make sure to set up your .env
file with the necessary API keys:
- COHERE_API_KEY
- JINA_API_KEY
- SERPER_API_KEY
- GROQ_API_KEY
We welcome contributions to OpenPerPlex! Please feel free to submit issues, fork the repository and send pull requests!
This project is licensed under the MIT License.
- Cohere for semantic chunking
- JINA AI for reranking
- serper.dev for Google search integration
- Groq for inference engine
- META opensource models
For any questions or feedback, please open an issue on this repository or contact me on X
Happy searching with OpenPerPlex! 🚀🔍