AIChat is an all-in-one command-line companion for large language models, written in Rust. One fast binary gives you a chat REPL, a shell assistant that turns plain English into commands, retrieval-augmented chat over your files, and function-calling agents — connected to more than twenty model providers. It has about 10,500 stars on GitHub and is actively maintained in 2026.
AIChat is built for people who live in the terminal. Switch freely between OpenAI, Claude, Gemini, Ollama, Groq, Mistral, local models and other providers with a single configuration, define roles to shape behavior for tasks like code review or translation, and pipe command output straight into a prompt. Its RAG mode indexes documents so you can chat with a codebase or knowledge base locally, while sessions, macros and a built-in REPL server round out scripting use. Because it is a single static binary with sensible keyboard-first design, it installs in seconds on Linux, macOS and Windows and stays out of the way until summoned. Configuration lives in a single file, and environments can be switched per shell or per project, which makes it easy to draft with cheap local models and escalate to frontier models for hard questions. AIChat also speaks the language of pipelines: it reads stdin, writes clean stdout, and offers a serve mode that exposes an OpenAI-compatible API for other tools. Its role system supports templating, so teams can share prompt presets through a git repository like any other configuration.
AIChat is completely free and open source under the Apache 2.0 and MIT licenses. You pay only your chosen model providers for API usage.
AIChat is for developers, sysadmins and terminal power users who want one fast CLI for every LLM task — chatting, shell help, and querying their own documents.