💰TechCrunch AI•Stalecollected in 46m
Google launches Antigravity 2.0 with new AI Ultra plan
💡New high-tier pricing and CLI tools for power users scaling AI workflows.
⚡ 30-Second TL;DR
What Changed
Launch of Antigravity 2.0
Why It Matters
Provides power users and developers with higher throughput and better tooling for intensive AI workflows.
What To Do Next
Evaluate your current API usage costs to see if the AI Ultra plan provides a better ROI for your high-volume tasks.
Who should care:Founders & Product Leaders
Key Points
- •Launch of Antigravity 2.0
- •New AI Ultra subscription tier at $100
- •5x usage limit increase over AI Pro
🧠 Deep Insight
Web-grounded analysis with 14 cited sources.
🔑 Enhanced Key Takeaways
- •Antigravity 2.0 is an AI-powered integrated development environment (IDE) that operates on an "agent-first" paradigm, enabling AI agents to autonomously plan, execute, and verify complex software tasks across various interfaces like code editors, terminals, and browsers.
- •The platform is a heavily modified fork of Visual Studio Code (VS Code) and was initially announced on November 18, 2025, alongside the release of Gemini 3, primarily leveraging Google's Gemini 3.1 Pro and Gemini 3 Flash models.
- •Antigravity 2.0 introduces dynamic subagents for parallelized workflows, scheduled tasks for background automation, and ecosystem integrations across Google AI Studio, Android, and Firebase.
- •The new AI Ultra plan, priced at $100/month, provides 5x higher usage limits in Google Antigravity compared to the AI Pro plan and includes access to the highest limits, prioritized traffic, and first access to new models, such as Gemini 3.5 Flash.
- •Beyond the desktop application, Antigravity 2.0 expands its ecosystem with an Antigravity CLI for terminal-based agent creation and an Antigravity SDK for programmatic access to define custom agent behaviors.
📊 Competitor Analysis▸ Show
Competitor Analysis: Google Antigravity 2.0 vs. AI Coding Tools
| Feature/Aspect | Google Antigravity 2.0 (AI Ultra) | GitHub Copilot (Pro+) | Claude Code (Max) | Cursor (Ultra) | OpenCode (OSS) |
|---|---|---|---|---|---|
| Core Paradigm | Agent-first IDE, autonomous task execution, orchestration | AI assistant, code completion, chat, agent mode | CLI-first agentic tool, deep logic, long context | AI-powered IDE, repo-level tooling, chat | Terminal agent, local-first, multi-model |
| Pricing | $100/month (AI Ultra) | $39/month (Pro+) | $100/month (Max) | $200/month (Ultra) | Free (software), pay for model API/local compute |
| Usage Limits | 5x Pro plan, highest limits, prioritized traffic, 25,000 AI credits/month | 50 premium requests/month (free tier), higher for Pro+ | Flat-rate, high usage (e.g., 10B tokens over 8 months for $100/month) | 20x Pro plan usage | Depends on model provider/local hardware |
| Primary Models | Gemini 3.5 Flash, Gemini 3.1 Pro, Gemini 3 Flash | GPT-5 mini, GPT-4.1, Opus 4.6 (Pro+) | Claude Sonnet 4.6, Claude Opus 4.6 | Various models, including Opus 4.6 | 75+ providers (Claude, GPT, Gemini, local) |
| Platform | Desktop app (Windows, macOS, Linux), CLI, SDK | IDE plugin (VS Code), CLI | CLI-first, IDE extensions | Standalone IDE (fork of VS Code) | Desktop beta, CLI (macOS, Linux, Windows) |
| Key Differentiator | Agent orchestration, verifiable "Artifacts," Manager View for parallel agents | Seamless integration with GitHub ecosystem | Predictable billing, strong reasoning, long context | Polished UX, large user base, powerful repo tooling | Privacy-first, multi-session, model-agnostic |
| Enterprise Support | Gemini Enterprise Agent Platform integration | Enterprise plans available | Enterprise plans available | Enterprise plans available | Open-source, flexible for custom enterprise integration |
🛠️ Technical Deep Dive
- Agent-First Architecture: Antigravity introduces an "agent-first" paradigm where AI agents operate with greater autonomy, capable of planning, coding, testing, orchestrating, and verifying work across the entire development stack.
- Fork of Visual Studio Code: The platform is a heavily modified fork of Microsoft's Visual Studio Code (VS Code), providing a familiar IDE interface.
- Core AI Models: Antigravity is primarily powered by Google's Gemini 3.1 Pro, Gemini 3 Flash, and now Gemini 3.5 Flash models. It also supports other AI models, including Anthropic's Claude Sonnet 4.6 and Claude Opus 4.6, as well as an open-source variant of OpenAI models (GPT-OSS-120B).
- Dual Views: It features an "Editor view" (a standard IDE interface with an agent sidebar) and a "Manager view," which acts as a control center for orchestrating multiple agents working in parallel across workspaces.
- Artifacts for Trust: To build user trust, agents generate "Artifacts"—verifiable deliverables such as task lists, implementation plans, screenshots, and browser recordings—rather than just raw tool calls.
- Agent Capabilities: Agents have direct access to the editor, terminal, and an integrated browser, and can learn from previous interactions. They can plan multi-step workflows, execute across different surfaces, and validate output.
- Sandboxing: For macOS users, agent terminal commands can be executed within a sandbox to prevent damage to files outside the workspace.
- Ecosystem Expansion: Antigravity 2.0 includes a standalone desktop application, a CLI tool for creating agents without a GUI, and an SDK for programmatic access and defining custom agent behaviors.
🔮 Future ImplicationsAI analysis grounded in cited sources
Accelerated Shift to Agentic Development
The launch of Antigravity 2.0 and the AI Ultra plan signifies Google's strong commitment to an agent-first development paradigm, potentially accelerating the industry-wide adoption of autonomous AI agents in software development.
Increased Developer Lock-in to Google's AI Ecosystem
By offering a comprehensive agent development suite, including an IDE, CLI, SDK, and integrations with other Google services, Google aims to position Antigravity as the default interface for building modern software, fostering deeper integration with its AI models and cloud services.
Enhanced Competition in AI-Powered IDEs
The new features and pricing tiers, especially the AI Ultra plan, will intensify competition among AI coding tool providers, pushing competitors to innovate further in agentic capabilities, usage limits, and model integration.
⏳ Timeline
2025-11-18
Google Antigravity 1.0 Public Preview Launch alongside Gemini 3
2026-01-13
Introduction of Agent Skills in Antigravity
2026-01-23
Terminal Sandboxing for macOS users introduced
2026-04-16
Antigravity 1.23.2 Preview Release
2026-05-19
Google Antigravity 2.0 and AI Ultra plan launch at Google I/O 2026
📎 Sources (14)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
📰
Weekly AI Recap
Read this week's curated digest of top AI events →
👉Related Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: TechCrunch AI ↗