IONOS Agent Docs v1.0 just launched!⭐ Star us on GitHub
docsOverview

IONOS Agent Starter Pack

A full‑stack intelligent agent powered by LangGraph ReAct architecture and IONOS AI Models Hub. It features dynamic web search, context-aware reasoning, and an extensible tool system for building production-ready AI assistants.

What’s inside

  • Backend: FastAPI service with a LangGraph ReAct agent (backend/).
  • Frontends: Next.js starter and Streamlit app (frontends/).
  • Docs: MDX documentation site (docs/).

Features

  • Chat with IONOS‑hosted models via a simple HTTP API.
  • Tools: real-time web search via Tavily API.
  • ReAct Agent: reasoning and acting with dynamic tool selection.
  • Easy to add your own tools and extend agent capabilities.

Requirements

  • Python 3.10+
  • Node.js 18+
  • IONOS_API_KEY (required)
  • TAVILY_API_KEY (required for web search functionality)

Quickstart

  1. Backend: follow the Windows‑friendly steps in Backend Setup.
  2. Choose a frontend:
  3. Send a chat request with header x-model-id (e.g., mistralai/Mistral-Small-24B-Instruct).

How it works

The client sends messages to FastAPI → the ReAct agent reasons about the request → tools are called dynamically as needed (e.g., web_search for real-time information) → the model produces a context-aware final answer.

Learn more:

Prerequisites