5ire

5ire (pronounced "fai-er") is a free, open-source desktop AI assistant and Model Context Protocol (MCP) client built with simplicity and user-friendliness at its core. It runs on macOS, Windows, and Linux, and connects to a wide range of major LLM service providers. The project is hosted on GitHub under a modified Apache 2.0 license and has accumulated over 5,000 stars.

What It Is

5ire is a cross-platform desktop application that acts as both a conversational AI assistant and an MCP client. It lets users bring their own API keys from providers like OpenAI, Anthropic, Google, Azure, Mistral, DeepSeek, Grok, Ollama, Baidu, Moonshot, and Doubao, then interact with those models through a polished local interface. The MCP integration means 5ire can connect to external tools and data sources using the standardized Model Context Protocol, extending conversations well beyond simple chat.

MCP Client and Tool Integration

One of 5ire's headline capabilities is its first-class support for MCP servers. MCP is described by the project as "a USB-C port for AI applications" — a standardized way to connect AI models to different data sources and tools. Through MCP servers, users can access the local file system, query databases, retrieve remote data, and run other tools directly from within the assistant. The project also maintains an open marketplace for MCP servers at mcpsvr.com, where users can discover and share server configurations. One-click server installation integration is documented for website operators who want to embed the capability.

Local Knowledge Base and RAG

5ire ships with a built-in local knowledge base powered by the bge-m3 multilingual embedding model. It supports parsing and vectorizing documents in docx, xlsx, pptx, pdf, txt, and csv formats, storing vectors locally to enable Retrieval-Augmented Generation (RAG) without sending data to external embedding services. This makes it suitable for users who want private, offline-capable document retrieval alongside their LLM conversations.

Productivity Features

Beyond MCP and RAG, 5ire includes several quality-of-life features:

• Prompts Library: Create and organize reusable prompts with variable support for flexible templating.
• Bookmarks: Save individual conversations or messages; bookmarked content persists even if the original messages are deleted.
• Global Search: Keyword search across all conversations to quickly locate past information.
• Usage Analytics: Track API usage and spending per provider to monitor costs and optimize usage.

Architecture and Setup

5ire is built primarily in TypeScript and uses native dependencies, which means it must be packaged per platform. macOS users can install via Homebrew (brew install --cask 5ire) or download Silicon/Intel binaries directly. Windows and Linux installers are also available. To use the MCP tools feature, Python, Node.js, and the uv Python package manager must be installed as they form the runtime environment for MCP servers. The project welcomes contributions and provides a Development Setup Guide for those who want to build on top of it.

Update: v0.15.4

The latest release is v0.15.4, published in March 2026. The repository shows active commit activity and ongoing issue resolution, signaling continued development momentum. The project's GitHub topics include mcp-client, model-context-protocol, knowledge-base, and llms, reflecting its dual identity as both a general AI assistant and a specialized MCP client.