Google Kills Gemini CLI and Replaces It With Antigravity CLI

Google just killed Gemini CLI. In its place: Antigravity CLI, a terminal agent that is not just a rename but a ground-up rethink of what a command-line AI tool should be. Announced at Google I/O 2026, it is the terminal surface of a four-product platform that also includes the Antigravity 2.0 desktop app, the Antigravity SDK, and the Antigravity IDE. The stakes are real: Gemini CLI and Gemini Code Assist IDE extensions will stop serving requests on June 18, 2026, for Google AI Pro and Ultra users, as well as free users of Gemini Code Assist for individuals.

From single-model chat to multi-agent orchestration

Gemini CLI proved the terminal could be an incredible interface for agentic tasks, but developer needs shifted. Teams now require multiple agents communicating with each other to split up work and solve complex problems, which means terminal tools need to share a unified backend with the rest of the workflow. That is the core architectural bet behind Antigravity CLI.

The tool is built in Go, supports asynchronous multi-agent workflows, and shares the same backend as the Antigravity 2.0 desktop application. This unified architecture means any future improvements to the core agents will instantly apply to the terminal experience. No more waiting for backports between surfaces.

The engine under the hood

By default, the CLI connects to Gemini 3.5 Flash (High), which was announced at Google I/O 2026. Google positions this model as running four times faster than other frontier models at the same tier, while outperforming Gemini 3.1 Pro on agentic and coding benchmarks. You can also swap models via the --model flag.

The key architectural concept here is dynamic subagents. In a single-agent setup, one model handles everything sequentially, and the context window fills up with the full history of every step, which degrades reasoning quality on later tasks. Subagents solve this by having an orchestrator agent receive your high-level goal, decide how to break it into subtasks, then spawn specialized subagents, each with its own isolated context window, running in parallel. The main agent can define and spawn specialized subagents on the fly to tackle focused subtasks in parallel, without polluting its own context window, and long-running operations run in the background so neither the main agent nor the UI is blocked.

What it can actually do

Antigravity CLI brings multi-step reasoning, multi-file editing, tool calling, and persistent history directly to your terminal, and is optimized for keyboard-driven workflows and remote SSH sessions. Here is what that looks like in practice:

• Parallel subagents: Spin up a data cleaner, analyzer, and visualizer simultaneously from a single prompt, with the orchestrator synthesizing the final output
• Slash commands: /goal runs a task to completion autonomously; /grill-me makes the agent ask clarifying questions first; /schedule sets up recurring or one-off invocations; /browser opts in to web access
• JSON hooks: These intercept and customize agent behavior at lifecycle stages (before a tool call, after a model call, at loop stop conditions) using simple JSON config files.
• Live voice transcription: Mic input transcribes in real time using Gemini Audio models, converting speech into clean, ready-to-send prompts.
• SSH-native: Detects remote sessions automatically and prints an authorization URL to complete login locally, no GUI required
• Antigravity plugins: Antigravity CLI keeps the most critical features of Gemini CLI: Agent Skills, Hooks, Subagents, and Extensions, now as Antigravity plugins.

Getting started in three commands

Installation is a single curl command. After that, you authenticate with a Google account or a GCP project, and the CLI identifies itself and connects to Gemini 3.5 Flash immediately.

# Install (macOS / Linux)
curl -fsSL https://antigravity.google/cli/install.sh | bash
# Launch
agy
# Run a multi-agent task to completion
/goal Analyze sales_raw.csv, clean it, and produce an HTML dashboard

If you were a Gemini CLI user with existing extensions, run agy plugin import gemini to migrate them. The official migration guide covers the full transition path, including how hooks map to the new JSON hook format.

Where it shines, and where it does not

Google's offerings should be especially enticing in light of the significant coding capability improvements of Gemini 3.5 Flash, as the model can perform many of the same tasks as frontier models at a fraction of the cost and output code far faster. Multiply those speed gains across several Antigravity agents, and ideas can come to life in minutes. The tool is a strong fit for large refactors you want to run in the background, data pipelines, remote server workflows over SSH, and CI/CD automation via the SDK.

The limitations are worth knowing upfront. Very custom build environments, obscure frameworks, or massive legacy codebases may still confuse the agent. You still need to understand architecture: Antigravity removes syntax friction, not design thinking. If you do not understand data flow, you will not recognize when the agent produces an inefficient schema. There is also a notable open-source gap: Gemini CLI had over 100,000 GitHub stars and 6,000 merged pull requests from hundreds of contributors. Antigravity CLI is closed-source, which has already drawn community criticism.

The bigger picture

Google used its I/O 2026 developer keynote to ship a meaningful architectural shift in how it packages AI-assisted development, announcing Google Antigravity 2.0 alongside an Antigravity CLI, an Antigravity SDK, Managed Agents in the Gemini API, and enterprise support through the Gemini Enterprise Agent Platform. The company is moving its developer tooling away from IDE-centric assistance and toward multi-agent workflow management as the primary abstraction.

The Managed Agents feature in the Gemini API is the most developer-friendly addition for teams that want to embed this capability in their own products. With a single API call, developers can spin up an agent that reasons, uses tools, and executes code in an isolated Linux environment. Managed Agents are powered by the Antigravity agent harness, built on Gemini 3.5 Flash, and available via the Interactions API and in Google AI Studio. Each interaction creates an environment that can be resumed in follow-up calls with all files and state intact, enabling seamless multi-turn sessions without reinitializing context.

The comparison is less about one command-line tool replacing another feature-for-feature and more about Google moving terminal AI work into a shared multi-agent system that spans terminal, editor, browser, and workspace-management surfaces. That is the real shift here. The terminal is no longer a standalone interface for a single model. It is one node in a coordinated agent network, and the Antigravity CLI GitHub repo makes clear that sessions can be exported to the desktop app and settings sync bidirectionally. Whether that architecture delivers on its promise in practice is the question the next few months will answer.