Common questions about the ultra-lightweight AI assistant
Find answers to your nano bot questions
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, the nano bot delivers core agent functionality in approximately 4,000 lines of Python code — 99% smaller than Clawdbot's 430,000+ lines. The nanobot ai supports multi-platform messaging, 11+ LLM providers, tool systems, memory management, and scheduled tasks.
The nanobot was developed by HKUDS, the Data Intelligence Lab at the University of Hong Kong. HKUDS is a research group focused on data intelligence and AI systems. The nano bot represents their work on ultra-lightweight AI architectures and research-friendly tools.
Yes! The nanobot ai is completely free and open source. The source code is available on GitHub under an open license. You can install the nano bot with pip install nanobot-ai at no cost. You only need to pay for LLM API access (Anthropic, OpenAI, etc.) which is separate from the nanobot itself.
The nano bot stands out through extreme simplification. With only ~4,000 lines of code (99% less than alternatives), the nanobot ai is dramatically easier to understand, modify, and deploy. It uses ~100MB RAM vs ~1GB for OpenClaw. The nanobot is research-friendly, allowing developers to comprehend the entire system in days instead of weeks.
The nanobot contains approximately 4,000 lines of core agent code. This is 99% smaller than Clawdbot (430,000+ lines) and OpenClaw (also 430,000+ lines). The nano bot proves that powerful AI assistants don't require massive codebases.
The nanobot was released on February 2, 2026. The current version is v0.1.3.post7 (released February 13, 2026). Despite being only 12 days old at the time of writing, the nano bot has already gained 17,800+ GitHub stars and 2,600+ forks.
Installing the nano bot is simple. Choose one of three methods: (1) PyPI (Recommended): pip install nanobot-ai, (2) uv Tool (Fast): uv tool install nanobot-ai, or (3) From Source (Latest): git clone https://github.com/HKUDS/nanobot && cd nanobot && pip install -e . See our complete installation guide for details.
The nanobot ai requires Python 3.10 or higher, approximately 100MB of RAM (estimated), and minimal disk space. It runs on Linux, macOS, and Windows — any operating system that supports Python 3.10+. You'll also need internet connectivity for LLM API calls and messaging platforms.
For most users, pip install nanobot-ai is recommended. This installs the latest stable version of the nano bot from PyPI. If you want faster installation, use uv tool install nanobot-ai. If you need the absolute latest development code or want to contribute to the nanobot, install from the GitHub source.
Yes! The nanobot runs on all major operating systems. Windows 10/11, macOS (Intel and Apple Silicon), and all Linux distributions are supported. The nano bot is pure Python, making it platform-independent. Just ensure you have Python 3.10+ installed.
Configure LLM providers by editing the nanobot ai configuration file (YAML or JSON format). Add your API keys for providers like Anthropic, OpenAI, DeepSeek, or Google Gemini. The nano bot supports 11+ providers. Example: Set llm_providers.anthropic.api_key to your Anthropic API key. See our configuration guide for detailed instructions.
The nanobot supports 8+ messaging platforms: Telegram (recommended), Discord, WhatsApp, Slack, Email, QQ, DingTalk (钉钉), Feishu (飞书), and Mochat. Each platform requires specific credentials. Telegram is the most feature-complete with voice transcription support.
To set up Telegram: (1) Message @BotFather on Telegram, (2) Create a new bot with /newbot, (3) Copy the bot token, (4) Add the token to your nano bot configuration file under platforms.telegram.bot_token, (5) Start the nanobot. Your Telegram bot will now respond to messages.
The nanobot ai memory system uses two components: (1) Long-term memory: Create a MEMORY.md file in your workspace for persistent knowledge, (2) Short-term memory: The nano bot automatically creates daily notes in memory/YYYY-MM-DD.md. Configure memory paths and search settings (70% vector + 30% BM25) in the configuration file.
The nanobot supports 11+ LLM providers: OpenRouter, Anthropic (Claude), OpenAI (GPT), DeepSeek, Google Gemini, Zhipu (智谱), DashScope, Moonshot, Groq, AiHubMix, and vLLM (for local models). The nano bot Provider Registry makes adding new providers a simple 2-step process.
The nanobot ai connects to Telegram, Discord, WhatsApp, Slack, Email, QQ, DingTalk (钉钉), Feishu (飞书), and Mochat. That's 8+ platforms total. The nano bot can manage all platforms simultaneously from a single instance with shared memory and context.
Yes! The nano bot has both long-term memory (MEMORY.md) and short-term memory (daily notes in memory/YYYY-MM-DD.md). The nanobot memory system was redesigned on February 12, 2026, for improved reliability. It uses hybrid semantic search (70% vector + 30% BM25) for efficient memory recall.
Yes! The nanobot ai includes a Cron system built on apscheduler for time-based automation. Schedule recurring tasks, set up automated workflows, and enable 24/7 autonomous operation. Example uses: hourly cryptocurrency monitoring, daily summary reports, periodic health checks.
The nano bot tool system includes file operations (read, write, create, delete), shell command execution, web access (fetch URLs, call APIs), and meta-capabilities. The nanobot architecture makes adding custom tools straightforward for developers.
Yes, the nanobot supports voice transcription on Telegram. Send voice messages to your Telegram bot and the nano bot will transcribe them before processing. This enables hands-free interaction with the nanobot ai.
The nano bot is inspired by OpenClaw but dramatically simplified. OpenClaw has 430,000+ lines and ~1GB RAM usage. The nanobot has ~4,000 lines (99% reduction) and ~100MB RAM (10x reduction). OpenClaw offers a full ecosystem with skills marketplace (3,000+ skills). The nanobot ai focuses on core features and research-friendliness. See our detailed comparison.
Clawdbot has 430,000+ lines of code and high resource usage. The nanobot has ~4,000 lines (99% less) and ~100MB RAM. Both are Python-based, but the nano bot achieves extreme simplification while retaining core AI assistant functionality. The nanobot is dramatically easier to understand and modify.
The nanobot ai achieves its small size through ruthless focus on essentials. HKUDS asked "What's truly necessary for an AI assistant?" and built only that. Clean architecture, no bloat, no redundancy. The nano bot proves that ~4,000 lines can deliver what others accomplish in 430,000+.
The nano bot has fewer features than OpenClaw but includes all core capabilities: chat sessions, multi-platform messaging, multi-LLM support, tools, memory, and scheduling. The nanobot lacks OpenClaw's skills marketplace and GUI, but for many use cases (research, learning, automation), the nano bot is fully capable.
Get help through several channels: (1) GitHub Discussions: Ask questions, (2) GitHub Issues: Report bugs, (3) DeepWiki: Read documentation. The nano bot community is active and helpful.
Report bugs on GitHub Issues. Provide detailed information: what you did, what happened, what you expected, error messages, and nanobot version. The HKUDS team actively monitors and responds to bug reports for the nano bot.
Yes! The nanobot ai welcomes community contributions. Submit pull requests for bug fixes, features, or improvements on GitHub. You can also contribute documentation, tutorials, or help other users. See the repository's CONTRIBUTING.md for guidelines.
The nano bot documentation is available in multiple places: (1) GitHub README: Quick start guide, (2) DeepWiki: Comprehensive docs, (3) This website: Guides and tutorials. All documentation is free and open.
The nanobot is written in Python. Specifically, Python 3.10+ is required. The nano bot leverages Python's rich ecosystem and ease of use while maintaining performance through efficient architecture. The nanobot ai codebase is clean, readable Python.
The nano bot uses approximately 100MB of RAM (estimated). This is 10x less than OpenClaw's ~1GB RAM usage. The nanobot ai is designed for minimal resource consumption, making it suitable for low-spec hardware and cloud environments.
Yes! The nanobot is designed for continuous operation. With its low memory footprint (~100MB) and stable architecture, the nano bot can run 24/7 without issues. Use the Cron system for scheduled tasks and automation. The nanobot ai is perfect for always-on deployments.
The nano bot requires internet for LLM API calls (unless using local models via vLLM) and messaging platforms. For fully offline operation, configure the nanobot with vLLM for local model support. The nanobot ai can then run without internet for inference, though platform messaging still needs connectivity.
The current nanobot version is v0.1.3.post7, released on February 13, 2026. Check GitHub Releases for the latest version and changelog. Update with pip install --upgrade nanobot-ai.
The nano bot is in early development (v0.1.3.post7) but functional. It's actively used for research, learning, and experimentation. For production use, understand that the nanobot ai is still evolving. The HKUDS team is rapidly adding features and improvements. Follow the GitHub repository for updates.
Didn't find your answer? Join the nano bot community on GitHub Discussions, read the comprehensive documentation, or explore our guides. The nanobot ai community is here to help.