Unveiled: Gemini CLI’s Latest Updates – Genkit, Jules, & Gemini 3 Pro, Revolutionize Coding!

Google continues to expand its AI toolkit for developers by upgrading Gemini CLI, its command-line interface dedicated to Gemini models. The tool has been enhanced with the addition of Gemini 3 Pro, a more interactive interface, and two significant extensions aimed at technical environments, Genkit and Jules.

Introducing Gemini 3 Pro in Gemini CLI

Google has now integrated Gemini 3 Pro directly into Gemini CLI. This update transforms the command line into a true technical assistant that supports complex tasks, from application generation to cloud performance analytics.

Access to Gemini 3 Pro is limited to users with a Google AI Ultra subscription or a paid API key. Once activated, the model can be invoked directly from the terminal to analyze code, generate complete projects, or orchestrate technical workflows without the need for a graphical interface.

Steps to access models in Gemini CLI:

An Enhanced Interactive Interface in the Terminal

Simultaneously, Google has evolved the interface of Gemini CLI by integrating support for interactive applications. The tool now relies on an internal virtual terminal capable of mimicking the behavior of a classic terminal, with dynamic screen refresh, cursor management, and full interface display.

This new technical layer allows users to manipulate editors like vim or nano, view tools such as htop, or use Git’s interactive screens without any display disruption, as Google points out, making the experience smoother and more in line with traditional terminal usage.

The improvements come with concrete demonstrations by the firm: generating complex applications from a simple description, transforming sketches into functional interface code, creating advanced shell commands from natural language instructions, automating documentation from existing code, or debugging cloud services by leveraging several tools within the same environment.

With this version, Gemini CLI transitions from a mere access point to an AI model to an interactive workspace driven by an advanced model. The tool integrates more closely with development workflows and paves the way for broader use of Gemini’s multimodal capabilities.

Two Extensions to Further Integrate AI: Genkit and Jules

Google has also enriched Gemini CLI with two major extensions designed to strengthen the integration of AI in technical environments: Genkit and Jules. These extensions are part of the recent evolutions of the CLI, aiming to transform the tool into a modular platform that aligns with developers’ practices.

Genkit: An Assistant That Understands Project Structure

The Genkit Extension enables Gemini CLI to seamlessly interact with projects based on Genkit, Google’s framework dedicated to AI applications. This extension allows the assistant to understand the project’s structure, identify defined workflows, navigate internal logic, or analyze generated traces.

The main benefit is contextualization. Instead of providing generic suggestions, Gemini can rely on the actual structure of the project to offer relevant recommendations. This facilitates prototyping, debugging, and exploration of applications built with Genkit.

Jules: An Autonomous Agent Working in the Background

The Jules extension offers a complementary approach focused on automation. It provides access to an agent capable of working in the background while the user continues to interact with Gemini CLI. Jules can analyze a repository, make corrections, restructure code, or propose modifications ready to be integrated into a project.

The aim is to delegate some repetitive or time-consuming operations to an autonomous agent to accelerate development cycles. Jules handles technical tasks while the user stays focused on their work in the terminal, further reinforcing the concept of a terminal “augmented” by AI.