Google unveils extension ecosystem for its command-line AI tool

On Wednesday, Google officially launched a major new feature for its command-line AI tool, Gemini CLI, introducing an extension system that allows third-party companies to integrate their services directly into the platform. Branded as “Gemini CLI Extensions,” the feature debuts with support from companies like Figma, Stripe, and others.

The move comes just two days after OpenAI rolled out in-app integrations for ChatGPT, which similarly allow external tools to plug into its AI environment. However, unlike ChatGPT’s tightly controlled app submission process, Gemini CLI Extensions are completely open — developers can publish their extensions without any approval or participation from Google. All available extensions will be publicly hosted on GitHub and can be manually installed by developers.

“This open ecosystem is absolutely essential for us,” said Taylor Mullen, a senior staff engineer on the project. “Everything we’re building is grounded in a fair and open environment where anyone can contribute.”

The first available extension is for Google’s own Nanobanana image generator, which was posted on GitHub last week. Once installed, it lets users generate images directly from the Gemini CLI terminal.

Since its launch in June, Gemini CLI has already surpassed one million users, according to Google, with adoption particularly strong among software developers. The tool is frequently used to develop and maintain codebases, something Google’s product managers are closely monitoring.

In an interview, Ryan J. Silva, Google’s Senior Director of Product Management for Developer Tools, said the new extension system is designed to transform Gemini CLI into “a fully extensible platform — a gateway for other tools and workflows that originate elsewhere in your toolchain.”