nanobot

nanobot is a self-hosted personal AI agent for people who want an assistant that can act on files and continue work across conversations. Its Python framework also suits developers building their own assistant integrations.

• Task execution: Read and edit files, run shell commands, search or fetch web pages, generate images and delegate work to subagents.
• Persistent work: Keep conversation history and long-term memory through Dream, pursue extended goals and run scheduled automations or tasks started by local triggers.
• Browser workbench: Organize project topics, switch models and workspaces, and inspect tool calls, file diffs, command output and generated artifacts; temporary chats stay out of saved history and memory.
• Tools and integrations: Extend the agent with MCP servers and reusable Skills, or connect other software through its Python SDK and OpenAI-compatible API.
• Chat access: Use the browser or terminal, or interact through Telegram, Discord, Slack, WeChat, email, Mattermost and Linear.

nanobot is open source under the MIT license. It can run locally or on your server, with Docker deployment and Linux and macOS service support. It supports local LLMs and OpenAI-compatible providers; choosing a cloud model sends model requests to that provider. Render can host the gateway and browser interface away from your machine. Scheduled delivery requires the gateway to remain running.