Antigravity Surfaces: Pick Your Weapon (Desktop, CLI, IDE, or SDK)
Quick answer
Antigravity offers four surfaces—Desktop, CLI, IDE, and SDK—all powered by the same agent harness. Find your perfect fit for managing autonomous agents.
So you’ve decided to dive into the Antigravity swamp. Good choice! But now you’re faced with a classic capybara conundrum: which surface should you use? Antigravity 2.0, CLI, IDE, or SDK? Each one is like a different water-weed patch in the same channel—they all connect to the same deep pool of agent power, but the way you wade in is up to you. Let’s break it down so you can pick the one that feels like home.
Quick Comparison
| Feature | Antigravity 2.0 | Antigravity CLI | Antigravity IDE | Antigravity SDK |
|---|---|---|---|---|
| Interface | Desktop App | Terminal (TUI) | Desktop App | Python Code |
| Best For | Multiple simultaneous tasks | Command-line / Headless | Directly editing code | Building custom agents |
The Four Surfaces
1. Antigravity 2.0
The default recommendation. This desktop app is like a calm capybara lounging in the sun while managing a whole fleet of agents. It lets you run multiple tasks across different projects without blocking your main workspace. You can switch between projects, monitor progress, and even schedule regular tasks like code quality checks or dependency updates. If you want a visual command center, this is your lily pad.
2. Antigravity CLI
For those who prefer the terminal—the true swamp dwellers. Built in Go for speed, the CLI is all about keyboard-driven navigation and headless execution. Need to run agents over SSH or inside a remote container? The CLI is your trusty paddle. It starts background agents without locking up your terminal, so you can keep slinging commands while your agents do the heavy lifting.
3. Antigravity IDE
If you’re the type who wants to see every line of code your agent touches, the IDE surface is for you. It puts agents right inside your workspace, letting you accept or reject changes line-by-line. Plus, with built-in debugging, the agent can spot runtime errors and offer one-click fixes. It’s like having a caiman assistant that doesn’t bite—unless you want it to.
4. Antigravity SDK (Python)
For the builders who want to craft their own custom agents from scratch. The SDK is a Python library that hooks into the same harness powering all Antigravity tools. You get direct access to the same tools and rules, and you can write an agent locally and deploy it to Google Cloud with zero code changes. Here’s a quick taste:
import asyncio
from google.antigravity import Agent, LocalAgentConfig
async def main():
config = LocalAgentConfig(
system_instructions="You are an expert assistant for codebase navigation.",
# api_key="your_api_key_here",
)
async with Agent(config) as agent:
response = await agent.chat("What files are in the current directory?")
print(await response.text())
asyncio.run(main())
Which One Should You Choose?
No wrong answers here. All surfaces run on the same agent harness, so you get the same core logic, plugins, and skills. Think of it like choosing between a Vercel-style deploy or a Supabase-style backend—different tools for different flows. Pick the one that matches your workflow: visual multitasker (2.0), terminal ninja (CLI), code whisperer (IDE), or agent architect (SDK).
For more details, check out the official docs and grab your download from the Antigravity Download Page.
Original announcement published on Google Cloud.