Configure Google Gemini CLI to Use KiosAPI Gateway

Set up the Google Gemini CLI to use KiosAPI as the backend provider, routing Gemini model requests through KiosAPI’s unified API gateway.

Google’s Gemini CLI is the official command-line interface for interacting with Gemini models. This tutorial walks you through installing Gemini CLI, configuring it to point at KiosAPI, and verifying that your setup works — so you can use Gemini models through KiosAPI’s unified gateway.

Prerequisites

• Node.js 18+ — verify with node --version
• A KiosAPI API key (starts with sk-) from the Dashboard → Token Management

Usage

Interactive mode

gemini

Switching Models

Change the default model in ~/.gemini/settings.json:

{
  "model": "gemini-2.5-flash"
}

Or specify a model for a single session:

gemini --model gemini-2.5-flash "Summarize this article"

Available Models

Model ID	Description
gemini-2.5-pro	Most capable Gemini model for complex reasoning
gemini-2.5-flash	Fast and cost-effective for everyday tasks
gemini-2.5-flash-lite	Ultra-fast lightweight model

Check the KiosAPI pricing page for the full list of available Gemini models and their costs.

Troubleshooting

Authentication failed

• Ensure you selected API Key (not Google Login) during setup
• Verify your API key is correct and starts with sk-
• Re-run gemini and re-select the auth method if needed

Connection error

• Check that GOOGLE_GEMINI_BASE_URL in ~/.gemini/.env is https://kiosapi.com/ — with a trailing slash
• Ensure base_url in settings.json also includes the trailing slash
• Verify your network connection

Model not found

• Check the model ID matches exactly (e.g., gemini-2.5-pro, not gemini-pro)
• Confirm the model is available on KiosAPI by checking the models list

Need Help?

• KiosAPI Dashboard
• Token Management Guide
• Available Models