This project utilizes the Google Generative AI API to create a conversational AI model. The model is based on the Gemini 2.0 Flash Exp model, a state-of-the-art language model capable of generating human-like text.
- New Chat button to start a new conversation.
- You can move between different conversation.
- Recent conversations are displayes in the sidebar.
- Completely same UI design.
- Node.js installed on your system
- A Google Cloud API key with the Generative AI API enabled
- Clone the repository to your local machine.
- Install the required dependencies by running
npm installin the project directory. - Replace
Your API Keywith your actual Google Cloud API key in thegemini.jsfile.
The conversational AI model can be configured using the generationConfig object. The following options are available:
- temperature: Controls the randomness of the generated text (default:
1) - topP: Controls the percentage of top tokens to consider when generating text (default:
0.95) - topK: Controls the number of top tokens to consider when generating text (default:
40) - maxOutputTokens: Controls the maximum number of tokens in the generated text (default:
8192) - responseMimeType: Controls the format of the generated text (default:
"text/plain")
- Building a chatbot for customer support
- Generating text for content creation
- Creating a conversational interface for a web application
The run function takes a single argument, prompt, which is the input text for the conversational AI model. The function returns a promise that resolves to the response from the model as a string.