How to Use the App

Streamlit Frontend

1. Go to http://localhost:8501
2. Choose Model Type:
   • Inference: IONOS Model Hub models with web search capabilities
   • Fine-tuned: Your custom IONOS Studio models
3. Select a model from the dropdown (models are fetched automatically)
4. Start chatting with the AI
5. For inference models, the agent will automatically search the web when needed
6. For fine-tuned models, you get specialized responses without web search

Features:

• Real-time streaming for inference models
• Conversation history
• Modern, responsive UI
• Instant message display

Next.js Frontend

1. Go to http://localhost:3000
2. Select an IONOS AI model from the dropdown
3. Start chatting with the ReAct agent
4. Experience real-time web search and context-aware responses

Note: The Next.js frontend is an alternative option and may not have all the features of the Streamlit app.

Both frontends handle conversation history locally and send the full message array to the backend for each request. The backend intelligently routes requests to either the ReAct agent (for inference models with web search) or directly to IONOS Studio (for fine-tuned models).