By HKUDS @ University of Hong Kong
Core agent functionality in ~4,000 lines — 99% smaller than Clawdbot
nanobot is an ultra-lightweight personal AI assistant developed by HKUDS (Data Intelligence Lab at the University of Hong Kong). Released on February 2, 2026, nanobot delivers core agent functionality in just ~4,000 lines of Python code — 99% smaller than Clawdbot's 430,000+ lines.
The nano bot architecture is clean, readable, and easy to understand. With only ~4,000 lines of core code compared to OpenClaw's massive codebase, nanobot ai is perfect for researchers and developers who want to study AI agent systems without drowning in complexity.
Installing nanobot is incredibly simple. Use pip install nanobot-ai or uv tool install nanobot-ai for stable releases, or clone from GitHub for the latest development version. The nanobot setup takes seconds, not hours.
The nanobot ai assistant integrates with 8+ messaging platforms including Telegram, Discord, WhatsApp, Feishu, Mochat, DingTalk, Slack, Email, and QQ. Run your nano bot agent across all your communication channels from a single lightweight installation.
nanobot supports OpenRouter, Anthropic (Claude), OpenAI (GPT), DeepSeek, Google Gemini, Zhipu, DashScope, Moonshot, Groq, AiHubMix, and vLLM for local models. Adding new providers requires just 2 steps thanks to the Provider Registry system.
Current Version: v0.1.3.post7 (Released February 13, 2026)
Language: Python 3.10+
License: Open Source
GitHub Stats: 17,800+ stars, 2,600+ forks
Despite its minimal code footprint, the nanobot framework delivers powerful AI agent capabilities that rival much larger systems.
Interactive conversations through multiple platforms with multi-LLM provider support. The nanobot ai handles context, memory, and tool use seamlessly across all supported chat platforms.
Connect nanobot to Telegram, Discord, WhatsApp, Feishu, Mochat, DingTalk, Slack, Email, and QQ. One nano bot instance manages all your messaging platforms with consistent behavior.
File operations, shell command execution, web access, and meta-capabilities. The nanobot tool system enables your AI assistant to interact with the real world through powerful, extensible tools.
Long-term memory stored in MEMORY.md plus short-term daily notes (memory/YYYY-MM-DD.md). The nanobot ai memory system was redesigned on February 12, 2026 for improved reliability with less code.
Time-based automation using apscheduler for job scheduling. Set up recurring tasks and scheduled messages that the nano bot processes automatically, enabling 24/7 autonomous operation.
Support for 11+ LLM providers including OpenRouter, Anthropic, OpenAI, DeepSeek, Google Gemini, and more. The nanobot Provider Registry makes adding new LLM providers a simple 2-step process.
Installing the nanobot ai assistant is quick and straightforward. Choose your preferred installation method below.
pip install nanobot-ai
The recommended method for most users. Get the latest stable release of nanobot from PyPI.
uv tool install nanobot-ai
Use the ultra-fast uv package manager for quick installation of the nano bot. Same stable release, faster installation.
git clone https://github.com/HKUDS/nanobot
cd nanobot
pip install -e .
Get the cutting-edge development version of nanobot directly from the GitHub repository.
nanobot is an ultra-lightweight personal AI assistant developed by HKUDS (Data Intelligence Lab at the University of Hong Kong). It delivers core agent functionality in just ~4,000 lines of Python code — 99% smaller than Clawdbot. The nano bot supports multiple messaging platforms, LLM providers, tools, memory management, and scheduled tasks.
The nanobot ai project was developed by HKUDS (Data Intelligence Lab @ HKU), a research group at the University of Hong Kong. It was released on February 2, 2026, and has rapidly gained popularity with 17,800+ GitHub stars and active community development.
nanobot contains approximately 4,000 lines of core agent code. This represents a 99% reduction compared to Clawdbot's 430,000+ lines. Despite the minimal codebase, the nano bot retains essential AI assistant functionality including tools, memory, multi-platform support, and LLM integration.
The nanobot ai assistant supports 8+ messaging platforms: Telegram (recommended), Discord, WhatsApp, Feishu, Mochat, DingTalk, Slack, Email, and QQ. You can run a single nano bot instance across multiple platforms simultaneously with consistent behavior.
Installing nanobot is simple with three options: (1) pip install nanobot-ai for stable releases from PyPI, (2) uv tool install nanobot-ai for fast installation with uv, or (3) clone from GitHub with git clone https://github.com/HKUDS/nanobot for the latest development version. All methods support Python 3.10+.
nanobot was inspired by OpenClaw but focuses on extreme simplicity and research-friendliness. While OpenClaw has 430,000+ lines of code and ~1GB RAM usage, the nano bot achieves core functionality with ~4,000 lines and ~100MB RAM. nanobot ai is ideal for researchers, minimalists, and those who want to understand AI agent architecture without excessive complexity.
Yes, nanobot is completely free and open source. The project is hosted on GitHub under an open source license. The nanobot ai community welcomes contributions through issues, pull requests, and discussions.
The nano bot requires Python 3.10 or higher, approximately 100MB of RAM (estimated), and any operating system that supports Python (Linux, macOS, Windows). You'll also need API access to at least one LLM provider and internet connectivity for cloud-based LLM services.
Join the growing community of developers and researchers using the nanobot ai assistant. With its minimal codebase, powerful features, and one-click deployment, getting started with nano bot has never been easier.