Google launches Antigravity 2.0 agentic dev suite

According to 9to5Google and the Antigravity product site, Google released Antigravity 2.0, a reworked agent-first development suite that centralizes local agent orchestration, project grouping, and parallel agent execution. Reporting by 9to5Google says the release includes a new Antigravity CLI that replaces the previous Gemini CLI, an Antigravity SDK for custom agents, and a Managed Agents feature in the Gemini API that provides an "isolated Linux environment" for agent execution. According to the Google AI Studio blog and 9to5Google, Google also expanded AI Studio's "vibe coding" features and launched a dedicated AI Studio Android app with export paths into local Antigravity. 9to5Google reports Google is adding a new $100 per month AI Ultra tier offering 5X usage limits for Antigravity workloads.
What happened
According to 9to5Google and the official Antigravity product site, Google released Antigravity 2.0, recasting Antigravity as an agentic development suite for running and managing multiple local agents in parallel. Reporting by 9to5Google documents new capabilities including project grouping, cross-workspace operations, scheduled messages, and local orchestration of multiple agents. According to 9to5Google, the release bundles an Antigravity CLI that replaces the previous Gemini CLI, an Antigravity SDK for prototyping custom agents, and a Managed Agents feature in the Gemini API that provides an "isolated Linux environment" for agent execution. 9to5Google also reports a new $100 per month AI Ultra tier that provides 5X the usage limits in Antigravity.
Technical details
According to the Google AI Studio blog post by Ammaar Reshi and Kat Kampf, AI Studio's updated "vibe coding" experience integrates the Google Antigravity coding agent to convert prompts into production-ready apps, with support for multiplayer features, secure API key storage, and Firebase backend integration. The Antigravity product site and Google I/O session overview describe the platform as an IDE and runtime combination offering a local, terminal-first workflow plus export/import links between AI Studio and local Antigravity projects. The Cloud blog recap frames Gemini 3 and related models as optimized for acting and coding, a capability that underpins the agentic functions highlighted across AI Studio and Antigravity.
Industry context
Editorial analysis: Companies are increasingly packaging agent orchestration, local development workflows, and managed runtimes together to shorten the loop from prototype to deployable agent. Observers in the developer tools space have documented a rising pattern where IDE-like surfaces couple with cloud-backed model runtimes and specialized CLIs to support both interactive and unattended agent workflows.
Practical implications for practitioners
Editorial analysis: For ML engineers and platform teams, the combination of a local agent-first IDE, a CLI, and a managed runtime raises operational questions around environment reproducibility, secrets management, and cost. Because 9to5Google reports Antigravity workloads are more token-intensive than chat use cases, teams will need to account for the reported new $100 AI Ultra tier or alternative quota strategies when evaluating long-running agent deployments.
What to watch next
Editorial analysis: Observers should track:
- •SDK and CLI compatibility across existing Gemini-based toolchains after the reported retirement of the Gemini CLI
- •details and limits of the Managed Agents runtime in the Gemini API
- •developer telemetry on whether the AI Studio <-> Antigravity export/import flows actually shorten iteration cycles in real projects. Public docs and codelabs referenced on Google's developer sites are the primary places to validate migration effort and runtime constraints
Bottom line
According to Google's documentation and reporting by 9to5Google, Antigravity 2.0 packages local agent orchestration, a terminal-first CLI, an SDK, and a managed execution option to support agentic development, while AI Studio's vibe-coding updates and a new Android app provide additional on-ramps for building and exporting projects into that agentic workflow. Editorial analysis: This line of product integration reflects an industry move toward toolchains built specifically for autonomous agent development rather than incremental extensions of chat-first tooling.
Scoring Rationale
This product update materially affects developer workflows for agentic applications by combining IDE, CLI, SDK, and managed runtimes. It is notable for practitioners building autonomous agents but not a frontier-model or regulatory milestone.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems