Four Must‑Try Open‑Source AI Tools: Gemini CLI, XiaoZhi Bot, AI Hub, GPT‑Pilot

1. Gemini CLI – Google’s Open‑Source AI Agent

Google recently open‑sourced Gemini CLI, which brings the Gemini large‑model directly to the command line. It supports Google Search, multimodal content generation, built‑in MCP, custom prompts and commands. The tool offers a generous free quota of 1,000 requests per day and 60 per minute, the highest among major AI services.

Open source address: https://github.com/google-gemini/gemini-cli

2. XiaoZhi – Voice‑Interactive Chat Robot

XiaoZhi is an ESP‑32‑based chatbot that understands spoken commands, leverages large models such as DeepSeek, and can control smart‑home devices via MCP services. The project has earned over 15 K stars on GitHub and supports more than 70 hardware boards with a detailed DIY guide.

Open source address: https://github.com/78/xiaozhi-esp32

3. AI Engineering Hub – Comprehensive LLM, RAG and Agent Tutorials

The AI Engineering Hub repository aggregates in‑depth tutorials and practical examples for building large‑model (LLM), Retrieval‑Augmented Generation (RAG) and AI‑agent applications. It covers models such as DeepSeek, Llama, and Gemma, and the material is compiled into a 500‑page PDF guide.

Open source address: https://github.com/patchy631/ai-engineering-hub

4. GPT‑Pilot – Open‑Source Programming Companion

GPT‑Pilot is an AI‑powered coding assistant that breaks large projects into small tasks, offers eight built‑in AI roles (product manager, architect, tester, etc.), and intelligently selects only the code relevant to the current task, addressing the challenge of long‑context codebases. The project has attracted more than 33 K stars on GitHub.

Open source address: https://github.com/Pythagora-io/gpt-pilot