lofcz/LLMTornado
The .NET library to build AI agents with 30+ built-in connectors.
What it solves
LLM Tornado solves the complexity of integrating various AI services into.NET applications. Instead of writing custom code for every different AI provider, developers can use a single, provider-agnostic SDK to interact with dozens of LLM APIs, local inference servers, and vector databases.
How it works
The toolkit uses a unified interface to communicate with over 30 API providers (like OpenAI, Anthropic, and Google) and local tools (like Ollama or vLLM). It employs an orchestration system for agents using three core concepts: an Orchestrator (the graph), a Runner (the node), and an Advancer (the edge). It also supports the Model Context Protocol (MCP) for connecting agents to external data and tools.
Who it’s for
It is designed for.NET developers building AI-powered applications, ranging from simple chatbots to complex, autonomous multi-agent workflows and enterprise-grade systems requiring guardrails and observability.
Highlights
- Multi-provider support: Built-in connectors for major LLM providers and local deployment tools.
- Agent Orchestration: Framework for coordinating specialist agents with support for handoffs and parallel execution.
- Multimodal capabilities: Supports text, images, video, audio, and documents.
- Extensive Integrations: Built-in connectors for popular vector databases like Chroma, Pinecone, and Qdrant.
- Enterprise Features: Includes support for guardrails, OpenTelemetry, and request transformation.