AI in your terminal. Run commands or ask questions - just hai it. Supports OpenAI, Claude, Gemini, DeepSeek & more.
# hai
> Say hi to AI, right from your terminal
[](https://www.npmjs.org/package/@aiany/hai)
[](https://www.gnu.org/licenses/gpl-3.0)
**hai** (pronounced "hi") brings AI to your terminal. Run commands or ask questions—just hai it.
```bash
$ hai how much disk space
# → runs `df -h`, explains the result
$ hai what is Docker
# → answers your question directly
```
No complex setup. No switching modes. No quotes needed. Just describe what you need.
## Features
- **Agent Mode** - AI automatically executes shell commands based on your intent
- **Multiple AI Providers** - OpenAI, Anthropic Claude, Google DeepSeek, and any OpenAI-compatible API
- **Profile Management** - Configure multiple profiles for different models and providers
- **Interactive Mode** - Multi-turn conversations in the terminal
- **Pipe Support** - Process input from stdin or files
- **Predefined Prompts** - Create reusable prompt templates with `{{input}}` placeholder
- **Thinking Mode** - Enable extended thinking for supported models (DeepSeek, o1/o3)
- **Streaming Output** - Real-time response streaming with color support
- **Keyboard Shortcuts** - Interrupt output with ESC or Ctrl+C
## Installation
```bash
npm install -g @aiany/hai
```
## Usage
### Basic
```bash
# Single question (quotes optional)
hai What is the capital of France
hai "What is the capital of France?"
# Interactive mode
hai -i
# Interactive mode with initial message
hai -i Hello
```
> **Note:** Shell special characters like `?`, `*`, `!` require quotes or escaping.
> Use `hai "who are you?"` or `hai who are you\?`
### Agent Mode
By default, hai runs in agent mode where AI can execute shell commands to help you:
```bash
# AI will run `ls -la` to answer
hai "What files are in this directory?"
# AI will check git status and logs
hai Show me recent changes in this project
# AI wilHAL 分层混合模型工作流 — 强模型(Claude)负责理解/拆解/验收,低成本模型(DeepSeek)负责检索/提取/清洗。Hermes Agent skill。
An LLM agent fine-tuned on DeepSeek for spaced repetition, dynamically integrating knowledge points based on the Ebbinghaus forgetting curve.
基于 STM32F103 构建的端到端 AI 智能手表生态。自研“零重定位”原生机器码动态加载引擎与页面栈式 UI 框架;集成生产级 OTA 回滚保护机制与高带宽(921600 baud)串口协议栈。通过 Node.js 中继实现 DeepSeek AI 语义控制及 ASRPRO 语音全双工交互,是一个集成了分布式计算、现代存储管理与 AI Agent 的嵌入式全栈工程。
A Meta-Agent-Driven Self-Evolving Multi-Agent System for UAV Detection and Tracking
One command to run Hermes AI Agent with a browser UI. Zero prerequisites. 一行命令,AI 就位。
网页应用Agent,接入DeepSeek、Mimo等模型