gemini-cli
SHA-256Google's open-source terminal AI agent for coding, debugging, and workflow automation powered by Gemini models.
Smart Download
Download Download Version
v0.65.0 · 36.3 MB
Chat with Google's Gemini in your terminal to write, debug, and automate code.
Core Features
- Query and edit large codebases with natural language, plus multimodal input from PDFs, images, and sketches
- Built-in file system, shell, and web fetch/search tools ready out of the box
- Extendable via MCP servers to connect GitHub, Slack, or Imagen/Veo media generation
- Scriptable headless mode with JSON and stream-json output for CI/CD automation
- Flexible auth: free Google account tier, Gemini API key, or enterprise Vertex AI
What It Can't Do
- •Google explicitly notes Gemini CLI is not optimized for the latest Gemini models (e.g., Gemini 4 Argon) and recommends Antigravity CLI for the best coding experience, so evaluate that migration path before adopting; the free tier has rate limits and heavy use requires a paid API key; Node.js is required, and WSL is recommended on Windows; token consumption can spike on very large repositories, so monitor usage to avoid burning through quota.
Use Cases
- Summarize yesterday's commits or explain an unfamiliar codebase's architecture
- Generate a working Discord bot from a provided FAQ document
- Run automated code tasks inside CI/CD pipelines using non-interactive mode
- Query open GitHub pull requests via MCP and post digests to Slack
Gemini CLI is Google's open-source AI agent that brings Gemini models directly into your terminal, offering a lightweight prompt-to-model workflow for coding, debugging, and automation. Compared with Claude Code, Gemini CLI is fully open source under Apache-2.0 and provides a generous free tier via Google account login, making it far cheaper to evaluate at scale; compared with Aider, it ships a richer built-in toolset—file system, shell, web fetch/search—plus native MCP server support for extensions like Imagen/Veo media generation. It supports multimodal input (generate apps from PDFs, images, or sketches), checkpointing, token caching, and a scriptable headless mode with JSON/stream-json output that fits neatly into CI pipelines. Authentication is flexible: personal Google account, Gemini API key, or enterprise-grade Vertex AI. One important caveat: Google states Gemini CLI is not optimized for the latest Gemini models (Gemini 4 Argon) and recommends Antigravity CLI for the best Gemini coding experience, so new users should weigh that transition before standardizing on it.
Tags
Getting Started
Install the software
Double-click the downloaded installer and follow the prompts
Install Node.js 20 or later
Run npm install -g @google/gemini-cli, or use brew install gemini-cli on macOS/Linux
Navigate to your project folder and launch it with the gemini command
- Install Node.js 20 or later
- Run npm install -g @google/gemini-cli, or use brew install gemini-cli on macOS/Linux
- Navigate to your project folder and launch it with the gemini command
SHA-256 checksum verified
Checksum extracted from GitHub official Release page
SHA256 Checksum
8c0f60a0726ddeacf83fa1c6edb725a9dfe0e001f6f6b3a036475e3d15fbf6ffThis checksum is extracted from the GitHub Release page. Verify file integrity after download.
All SHA-256 checksums on this platform are extracted from the project's official GitHub Release page, without any modification. You can independently verify them on the GitHub Releases page.
Open Source Transparency
View GitHub SourceUninstall Info
If installed via npm, run npm uninstall -g @google/gemini-cli; if installed via Homebrew, run brew uninstall gemini-cli. To remove all settings and cached data, delete the ~/.gemini directory in your home folder.
No Extra Dependencies
Ready to use after download. No additional runtime required.
Having issues? Check the FAQ below
4 FAQs
Similar Projects
mempalace
Open-source local-first AI memory system with verbatim storage and semantic search — zero API calls, no keys, top benchmark scores.
ollama
Ollama lets you download, run, and manage large language models locally. One command, multiple platforms, endless possibilities.
ComfyUI
ComfyUI is a modular AI engine with a node-based graph interface for generating images, video, 3D models, and audio. Unlike traditional webui tools, it gives pixel-level control over every model and parameter, making it the preferred choice for professionals.