Argo

Argo is an open source desktop AI agent platform that runs on the user's own machine. It lets people build and use assistants that plan tasks, call tools and work through complex requests such as deep research, while keeping data stored locally. It suits individuals and teams who want a private, local assistant rather than a cloud chat service. Models can be downloaded and run through Ollama with one click, or connected through any LLM API compatible with the OpenAI format. An agent factory lets users describe a specialist, such as a travel planner, industry analyst or legal advisor, in plain language and configure it with minimal setup. A multi agent architecture handles deep research by planning steps and calling tools autonomously. The local RAG knowledge base ingests files, folders and websites with dynamic folder synchronization, and an agentic mode breaks complex questions into parts for fuller answers. MCP support comes with built in tools for browser control and web scraping. Argo is free to download for Windows 10 and 11 and for macOS on Intel and Apple silicon, and can also run under Docker. The source is published on GitHub. The vendor does not publish any paid plans or pricing. The project's repository describes Argo as bringing a local Manus style agent to the desktop, and data stays on the user's device, which gives control over privacy and data sovereignty.