A simple web chat application supporting popular LLM API interfaces.
Based on @xtekky's chatgpt-clone project, this project aims to provide a unified web platform for easily switching between different LLM models, such as LLaMa-3, DeepSeek, ChatGPT, Mixtral, Gemma etc. You can easily add more LLM API models by modifying config.json.
You can also visit https://miaomiao88.top for a demo.
- Clone this project:
git clone https://github.com/pisces76/chatgpt-llms.git; cd chatgpt-llms
- Create a virtual environment and install dependencies:
python -m venv venv
source venv/bin/activate # On Windows, use `venv\Scripts\activate`
pip install -r requirements.txt
- Copy config.json.example to config.json
- Modify the models section in config.json with your correct API KEY, for example:
"models": [
{
"llm_name": "deepseek",
"api_key": "sk-xxxx",
"api_base": "https://api.deepseek.com",
"model": "deepseek-chat",
"developer": "DEEPSEEK"
},
...
]
- You can add more models in config.json
- For testing, use Python Flask:
python run.py
- For production, running start.sh script to startup gunicorn:
./start.sh &
- Now you can visit http://ip:8080 for a chat.
This is my first Github code project. If you find this project useful and would like to support it, please consider making a donation. Your support will help to ensure the continuity and development of the project.