Unlock AI Power in Your Terminal: A Complete Guide to Gemini CLI

What is Gemini CLI?

Gemini CLI is a command‑line AI workflow tool released by Google that brings Gemini models (e.g., Gemini 2.5 Pro) directly into the terminal.

It can be invoked with natural language to perform coding, debugging, content generation, research, and task management.

Key Features

No tool switching: Write code and solve problems directly in the terminal without opening an IDE or web page.

Huge context window: Supports up to 1 million tokens, allowing processing of large codebases and documents.

Rich out‑of‑the‑box functions: Besides code, it can draft copy, look up information, manage pipelines, generate media, essentially acting as a full‑featured AI assistant in the shell.

Generous free quota: A personal Google account gets ~60 requests per minute and 1 000 requests per day, one of the most generous preview plans.

Open source: Released under the Apache 2.0 license, welcoming community contributions and audits.

Installation

Prerequisite: Node.js 18 or newer. node -v Install via npx: npx https://github.com/google-gemini/gemini-cli Or via npm:

npm install -g @google/gemini-cli

Configuration

After installation run gemini to start the interactive CLI. The first run may request permissions.

Login with a Google account (free 1 000 daily requests) or set a Gemini API key: export GEMINI_API_KEY="your_key" If you are behind a proxy, set HTTP/HTTPS proxy environment variables, e.g.:

export HTTP_PROXY="http://127.0.0.1:7890"
export HTTPS_PROXY="http://127.0.0.1:7890"

For Google Cloud projects, set the project ID:

export GOOGLE_CLOUD_PROJECT="your_project_id"

Using with VS Code

Create a new project, open it in VS Code, configure the proxy if needed, then run gemini in the integrated terminal. Example prompt: gemini > Write a 2048 game The CLI will generate the source files, ask for permission to create them, and display progress. After completion the generated files appear in the explorer.