Building the agentic future: Developer highlights from I/O 2026

At Google I/O 2026, Google is accelerating the shift from “text prompts” to real execution with the launch of Gemini 3.5 Flash. This model combines advanced intelligence with extreme speed, outperforming Gemini 3.1 Pro across most benchmarks while running four times faster than other leading models, delivering the performance required for real-world agentic workflows.

Google is putting these capabilities directly into developers’ hands with the launch of the new Google Antigravity 2.0 desktop application, Managed Agents in the Gemini API, and native Android vibe coding support inside Google AI Studio. Below are the key announcements for developers building the agentic future.

The Google Antigravity ecosystem

Antigravity is Google’s agent-first development platform that enables developers to turn ideas into production-ready applications. Google is expanding the ecosystem to manage and deploy agents that integrate across major development surfaces.

Here are the new ways to work with Antigravity:

• Antigravity 2.0: A new standalone desktop application that delivers a fully agent-optimized experience. It serves as a central hub for agent interaction, allowing developers to orchestrate multiple agents in parallel. It also introduces dynamic sub-agents for parallel workflows, scheduled tasks for background automation, and deeper integrations across Google AI Studio, Android, and Firebase.
• Antigravity CLI: For developers who prefer the terminal, Antigravity CLI provides a lightweight, high-speed interface for creating new agents instantly without a graphical UI. Google encourages Gemini CLI users to migrate to Antigravity CLI.
• Antigravity SDK: The SDK provides programmatic access to the same underlying agent infrastructure powering Google’s products. Optimized for Gemini models, it enables developers to define custom agent behaviors and deploy them on their preferred infrastructure.
•  
• Antigravity in the Gemini Enterprise Agent Platform: This integration simplifies enterprise workloads by allowing Google Cloud customers to connect Antigravity directly to their Google Cloud projects.
• Google AI Ultra subscription, starting at $100: Google is introducing a new $100/month AI Ultra plan designed to bring frontier AI into developer workflows, offering a 5x higher usage limit in Google Antigravity compared to the AI Pro plan. For a limited time, new and existing subscribers will receive $100 in bonus Antigravity credits, activated when plan quotas are reached. The offer can be claimed in the Antigravity app and expires on May 25, 2026, with terms applying.

Managed agents in the Gemini API

Google is introducing Managed Agents in the Gemini API. With a single API call, developers can now spin up an agent capable of reasoning, using tools, and executing code inside an isolated Linux environment.

These agents are powered by the Antigravity agent harness, built on Gemini 3.5 Flash, and are available through the Interactions API and Google AI Studio. Key features include:

• Antigravity agent harness: Developers gain access to the same infrastructure used to power Google’s internal agents. The system is co-optimized with Gemini models, especially Gemini 3.5 Flash.
• Persistent, isolated environments: Each interaction creates an environment that can be resumed in later calls, preserving all files and state for seamless multi-turn workflows.
• Custom agent definitions: Developers can extend Antigravity agents with custom instructions and skills using Markdown files, and quickly start from new agent templates in the Google AI Studio Playground.

Google AI Studio mobile and ecosystem integrations

Google AI Studio, powered by the Antigravity coding agent, enables developers to turn ideas into applications using the latest Gemini models. Google is expanding where and how it can be used:

• Google AI Studio mobile app: Available for pre-registration this week, the new mobile app allows developers to capture ideas on the go and generate working prototypes by the time they return to their desk.
• Workspace integration: Agents can now natively call Google Workspace APIs and embed functionality directly into applications.
• Export to Antigravity: A new integration allows entire projects to be transferred from Google AI Studio into local Antigravity development—or production—with a single click, preserving full project context so developers can continue seamlessly.
• Native Android support: Developers can now build high-quality Android apps from a single prompt. Google is also adding Google Play Console support inside AI Studio, enabling direct publishing to testing tracks.

The Build with Gemini XPRIZE Hackathon

Google is launching the Build with Gemini XPRIZE Hackathon, a global competition with a $2 million prize pool—the largest in hackathon history. Developers can build using Gemini models, tools, and platforms to create real-world applications addressing major global challenges, from reducing food waste to advancing scientific research. Finalists will present their projects live at the Moonshot Gathering in Los Angeles this September, competing for the grand prize.

The shift from single prompts to always-on collaborative agents is fundamentally changing how software is built—and this is only the beginning. Google’s goal is to make this transition seamless for developers. With the new Antigravity platform, Managed Agents in the Gemini API, and expanded Google AI Studio capabilities, Google aims to remove infrastructure friction so developers can focus entirely on bringing their ideas to life.