Google Unveils Gemini 3 Pro and Debuts Antigravity Agentic Platform

November 18, 2025

Details

  • Google has launched Gemini 3 Pro, its most advanced AI model to date, now available in public preview as of November 18, 2025.
  • Gemini 3 Pro can be accessed via Google AI Studio, Vertex AI, the Gemini API, and the newly introduced Antigravity platform, which serves as an agentic development environment compatible with Mac, Windows, and Linux.
  • The model features state-of-the-art multimodal understanding, advanced spatial reasoning, a one-million-token context window, video comprehension capabilities, and introduces "vibe coding," which enables developers to build interactive applications from natural language prompts.
  • Antigravity empowers AI agents to autonomously navigate coding, terminal, and browser interfaces, streamlining complex, end-to-end software development tasks and surpassing the automation abilities of Gemini 2.5 Pro.
  • Pricing is set at $2 per million input tokens and $12 per million output tokens for up to 200,000 tokens, with integrations extending to third-party tools like Cursor, GitHub, JetBrains, Android Studio, and Replit.

Impact

Google's latest release sharpens its competitive edge against industry leaders like OpenAI and Anthropic, especially in agentic development and multimodal AI. The adoption of "vibe coding" and autonomous agent workflows signals a shift from code-completion tools toward AI as strategic development partners, which is poised to significantly accelerate enterprise AI adoption and reshape developer workflows industry-wide.

Read more