What is PicoClaw?
PicoClaw is an ultra-lightweight personal AI assistant written in Go. It deploys as a single binary with less than 10MB memory usage, supports 16+ chat platforms (Telegram, Discord, Slack, QQ, DingTalk, etc.), and runs on everything from a $10 Raspberry Pi to cloud servers.
How is PicoClaw different from ChatGPT or other AI assistants?
PicoClaw is self-hosted — your data stays on your device. It connects to LLM providers (OpenAI, Claude, DeepSeek, etc.) via API, but all conversation history and configuration remain local. It's also designed for ultra-low resource environments that cloud-based solutions can't serve.
What hardware do I need to run PicoClaw?
Minimum: any device with 64MB RAM and an internet connection for LLM API calls. Recommended: 512MB RAM. PicoClaw runs on x86_64, ARM64, ARMv7, ARMv6, RISC-V, and MIPS architectures. A $10 Raspberry Pi Zero works perfectly.
Which LLM providers does PicoClaw support?
OpenAI (GPT-4o, o1), Anthropic (Claude), DeepSeek, Google Gemini, Mistral, Volcengine (Doubao), Qwen, Minimax, Moonshot (Kimi), Cerebras, Perplexity, Ollama (local), and any OpenAI-compatible API.
Can I run PicoClaw on my phone?
Yes! On Android, just download and install the APK — no root or special setup needed. You can also run PicoClaw via Termux if you prefer the command-line approach. See the Android guide for details.
Is PicoClaw free and open source?
Yes, PicoClaw is fully open source under the MIT-compatible license. The software is free. You only pay for the LLM API usage (e.g., OpenAI, Claude) based on your own consumption.