Google has introduced Antigravity 2.0, a major update to its AI coding and agent orchestration platform, expanding its push into agentic software development and automated coding workflows.
Announced at Google I/O 2026, the new release adds a redesigned desktop application, a dedicated command-line interface tool, and an SDK that allows developers to build custom AI agents and workflow automations on top of Google’s coding infrastructure.
Google originally launched Antigravity last year as its response to the growing market for AI-native coding platforms such as Cursor, Claude Code, and Replit. The new version significantly expands the platform’s scope beyond code generation into orchestration of long-running autonomous workflows.
According to Google, the updated desktop application can coordinate multiple AI agents simultaneously, allowing users to execute tasks in parallel, schedule background operations, and design custom subagent workflows. The platform also integrates directly with Google AI Studio, Android development environments, and Firebase services.
Much of the system is powered by Google’s new Gemini 3.5 Flash model, which the company said was partially developed using Antigravity itself.
Google is additionally adding native voice interaction to Antigravity, allowing developers to issue spoken commands inside the coding environment. The feature mirrors Google’s broader rollout of voice-driven AI interactions across products including Gmail, Docs, and Search.
Google Pushes Deeper Into Agentic Coding
The release reflects the accelerating shift across the software industry toward AI agents capable of handling increasingly autonomous development tasks.
Alongside the desktop platform, Google launched a new Antigravity CLI tool aimed at developers working primarily in terminal environments. The company is encouraging users of its earlier Gemini CLI tooling to migrate to the new system.
Google also introduced an Antigravity SDK that enables developers and enterprise customers to build custom agents and workflow systems on top of the platform. Google Cloud customers will be able to connect enterprise infrastructure directly into Antigravity environments, while AI Studio will provide templates for building custom enterprise agents.
The company additionally launched an export tool inside AI Studio that allows developers to move projects from browser-based workflows into local development environments for further customization.
Google said Antigravity’s capabilities are also beginning to appear inside consumer-facing products. In Search, for example, users will increasingly see dynamically generated interfaces and lightweight applications created in real time as part of AI-generated responses.
Premium AI Developer Tools Become a Revenue Layer
Google also introduced new pricing changes tied to its AI subscription business as AI coding platforms become increasingly resource-intensive.
The company launched a new AI Ultra subscription tier priced at $100 per month that provides five times higher Antigravity usage limits than the standard Pro plan. At the same time, Google reduced the price of its highest-end AI Ultra plan from $250 to $200 per month while maintaining higher compute limits.
The pricing changes mirror broader moves across the AI industry, where companies including OpenAI and Anthropic have introduced premium subscription tiers targeting developers, enterprises, and power users running large-scale AI workflows.
The launch also builds on Google’s wider Gemini expansion announced at I/O, including Gemini 3.5 Flash for coding and agentic workflows, Gemini Omni Flash for multimodal video generation, and new AI-powered Android app creation tools inside Google AI Studio. Together, the products reflect Google’s effort to position Gemini and Antigravity as a unified ecosystem for software development, automation, AI agents, and multimodal content generation.
