Llama MCP Streamlit

What is Llama MCP Streamlit?

Llama MCP Streamlit is an interactive AI assistant built with Streamlit, NVIDIA NIM's API (LLaMa 3.3:70b)/Ollama, and Model Control Protocol (MCP). It provides a conversational interface where users can interact with an LLM to execute real-time external tools via MCP, retrieve data, and perform actions. The app runs as a Streamlit web application and is intended for developers and users who want an LLM-powered chat with tool integration.

How to use Llama MCP Streamlit?

1. Configure the .env file with your API keys for NVIDIA NIM or Ollama.
2. Install dependencies using Poetry: poetry install.
3. Run the app with poetry run streamlit run llama_mcp_streamlit/main.py.
   Alternatively, use Docker: docker build -t llama-mcp-assistant . then docker compose up.
4. To change the MCP server, edit utils/mcp_server.py and set the desired StdioServerParameters (NPX or Docker).

Key features of Llama MCP Streamlit

• Real-time tool execution via MCP
• LLM-powered chat interface with Streamlit
• Support for multiple LLM backends (NVIDIA NIM & Ollama)
• Custom model selection and API configuration
• Docker support for easy deployment
• Configurable MCP server using NPX or Docker

Use cases of Llama MCP Streamlit

• Interact with an LLM that can execute external tools to retrieve or modify data
• Build a conversational AI assistant that accesses files and performs actions
• Experiment with MCP tool integration in a Streamlit-based chat interface
• Use different LLM backends (NVIDIA NIM or Ollama) for the same tool-driven workflow

FAQ from Llama MCP Streamlit

What LLM backends are supported?

The app supports NVIDIA NIM's API (LLaMa 3.3:70b) and Ollama, configured via environment variables.

How do I configure the MCP server?

Update the utils/mcp_server.py file to set StdioServerParameters for either NPX (e.g., @modelcontextprotocol/server-filesystem) or Docker (e.g., mcp/filesystem).

What is the license for Llama MCP Streamlit?

The project is licensed under the MIT License.

Can I run the app without Docker?

Yes, the app can be run using Poetry (Python 3.11+ required) – see the installation steps above.