TechHelpHub
中文
AD

Sponsor Us|If you find this site helpful, please consider sponsoring

Support Now →
• tutorial

Gemini CLI Installation Guide

Complete installation and configuration for Google Gemini CLI

#AI #Gemini #Google

Gemini CLI Installation Guide

Gemini CLI is a command-line tool provided by Google that allows you to use the Gemini large language model in your terminal.

Environment Preparation

  • Go 1.21 or higher (optional, for source compilation)
  • Valid Google Cloud account
  • Gemini API access

Installation Methods

# macOS
brew install gemini-cli

Linux (Snap)


sudo snap install gemini-cli


Windows (Chocolatey)


choco install gemini-cli

Method 2: Using npm

npm install -g @google/gemini-cli

Verify installation


gemini —version

Method 3: Source Compilation

git clone https://github.com/google/gemini-cli.git
cd gemini-cli
go build -o gemini
sudo mv gemini /usr/local/bin/

Authentication Configuration

1. Get API Key

Visit Google AI Studio to create an API Key.

2. Configure Credentials

# Method 1: Interactive configuration
gemini config

Method 2: Direct setup


gemini config set api_key YOUR_API_KEY


Method 3: Environment variables


export GEMINI_API_KEY=“YOUR_API_KEY”
export GOOGLE_CLOUD_PROJECT=“your-project-id”

Common Commands

Basic Chat

# Interactive mode
gemini chat

Single query


gemini ask “Explain the quicksort algorithm”

# Code generation
gemini code generate "Python HTTP server"

Code review


gemini code review ./src/main.go


Unit test generation


gemini test generate ./utils.py

File Processing

# Analyze file
gemini analyze ./document.pdf

Multi-file context


gemini chat —context file1.py,file2.js

Advanced Configuration

# ~/.gemini/config.yaml
model: gemini-pro
temperature: 0.7
max_output_tokens: 2048
top_p: 0.95
timeout: 60
retry_attempts: 3

Practical Tips

  1. Use aliases: Add alias g='gemini chat' to .bashrc or .zshrc
  2. Pipe input: cat file.py | gemini explain
  3. Output formatting: gemini ask --format json "API design"