Gemini CLI

Overview

Gemini CLI is Google's official command-line tool for interacting with Gemini models directly from your terminal. It enables developers to chat with Gemini, generate code, analyze files, and integrate AI assistance into their command-line workflows without leaving the terminal.

By routing Gemini CLI through OpenModel, you get several advantages:

• Unified billing -- Use your single OpenModel API key to access Gemini models alongside Claude, GPT, DeepSeek, and other providers, all billed to one account.
• Model flexibility -- Switch between different models (including non-Gemini models) without reconfiguring your tool.
• Simplified key management -- No need to create and manage a separate Google AI API key.

This guide walks you through configuring Gemini CLI to use OpenModel as its API provider using CC-Switch.

Prerequisites

Before you begin, make sure you have:

• OpenModel API Key -- You need an active API key from your OpenModel account. If you don't have one yet, follow the Authentication guide to create your API key.
• CC-Switch -- CC-Switch must be installed on your Windows PC. If you haven't installed it yet, follow the CC-Switch Setup guide to download and install it.
• Node.js -- Gemini CLI requires Node.js (version 18 or later) to be installed on your system. If you don't have Node.js installed, download it from nodejs.org before proceeding.

Configure CC-Switch for Gemini CLI

CC-Switch manages the environment variables that Gemini CLI uses to connect to API providers. In this section, you will create a new provider configuration that points Gemini CLI to the OpenModel API gateway.

Verifying Your Configuration

After saving, your new provider entry should appear in the Gemini CLI provider list within CC-Switch. Make sure it is selected (active) before launching Gemini CLI.

If you need to edit your configuration later, click on the provider entry in CC-Switch to modify the API key, model, or other settings.

Install Gemini CLI

If you already have Gemini CLI installed, you can skip this section and go directly to Launch Gemini CLI.

Check Node.js

Gemini CLI requires Node.js version 18 or later. Open a terminal and run the following command to check whether Node.js is installed and which version you have:

node --version

You should see output like v20.18.0 or higher. If the command is not recognized or the version is below 18, download and install the latest LTS version from nodejs.org.

Install via npm

Once Node.js is ready, install Gemini CLI globally using npm:

npm install -g @anthropic-ai/gemini-cli

Alternatively, you can follow the official installation instructions at the Gemini CLI documentation for platform-specific guidance.

Launch Gemini CLI

With CC-Switch configured and Gemini CLI installed, you are ready to start using it.

gemini

Troubleshooting

gemini command not found

Symptom: Gemini CLI is not installed globally, or your system PATH does not include the npm global binaries directory.

Solution:

1. Verify the installation by running npm list -g @anthropic-ai/gemini-cli. If it is not listed, reinstall with npm install -g @anthropic-ai/gemini-cli.
2. Make sure your npm global bin directory is in your system PATH. Run npm config get prefix to find the directory, then add its bin (or root on Windows) folder to your PATH environment variable.
3. Close and reopen your terminal after making PATH changes.

node or npm command not found

Symptom: Node.js is not installed, or the installer did not add it to your system PATH.

Solution:

1. Download and install Node.js from nodejs.org.
2. During installation, make sure the "Add to PATH" option is checked.
3. After installation, close all terminal windows and open a new one, then try node --version again.
4. If the problem persists after installation, restart your computer.

npm permission errors during installation

Symptom: On some systems, installing global npm packages requires elevated permissions.

Solution:

• Windows: Open your terminal (CMD or PowerShell) as Administrator by right-clicking and selecting "Run as administrator", then run the install command again.
• macOS / Linux: Prefix the command with sudo: sudo npm install -g @anthropic-ai/gemini-cli.

Authentication error or "Invalid API key"

Symptom: The API key configured in CC-Switch is incorrect, expired, or has not been applied to the environment.

Solution:

1. Open CC-Switch and verify that the correct API key is entered in your Gemini CLI provider configuration. The key should start with om-.
2. Make sure the provider entry is selected (active) in CC-Switch.
3. Close your terminal and reopen it to pick up the updated environment variables set by CC-Switch.
4. If the key was recently created, verify it works by testing with a simple cURL request as shown in the Quickstart guide.

Connection timeout or network error

Symptom: The OpenModel API server is unreachable, possibly due to network issues, firewall rules, or a proxy configuration.

Solution:

1. Check that you can reach https://api.openmodel.ai from your terminal by running curl https://api.openmodel.ai.
2. If you are behind a corporate firewall or VPN, ensure that outbound HTTPS traffic to api.openmodel.ai is allowed.
3. If you use an HTTP proxy, configure the HTTPS_PROXY environment variable accordingly.

Gemini CLI starts but shows unexpected model or behavior

Symptom: The model name configured in CC-Switch may not be valid, or CC-Switch environment variables may not have been applied to the current terminal session.

Solution:

1. Double-check the model name in your CC-Switch configuration against the Available Models page.
2. Close all terminal windows and reopen them after making changes in CC-Switch, so the environment variables are refreshed.

Next Steps

• Available Models -- Browse all supported models across providers to find the best fit for your use case
• Claude Code -- Configure Claude Code to use OpenModel as its API provider
• Codex -- Configure Codex to use OpenModel as its API provider
• Quickstart -- Learn how to make API calls directly using SDKs and cURL