# LLMQore

A Qt and C++ library that gives a desktop application cloud and local LLM providers, MCP clients and servers, and ACP agent hosting, with streaming delivered as ordinary Qt signals.

LLMQore is a Qt and C++ library for connecting a desktop application to cloud and local LLM providers, MCP clients and servers, and ACP agents. It talks to Claude, OpenAI, Google AI, Mistral, DeepSeek, Qwen, Ollama and llama.cpp, delivering streaming output as Qt signals on the object's own thread and async results as QFuture, so it can connect directly and safely into a widget or model.

Tool calling exposes a base tool class to the model with a built-in tool loop, and an MCP client and server share the same tool registry over stdio, Streamable HTTP or SSE, with a bridge CLI that puts several upstream MCP servers behind one endpoint. Conversation history is translated into each provider's own shape so it stays portable between them. It is MIT licensed.

Website: https://github.com/Palm1r/llmqore