Start building with Gemini 3 Pro

Google is launching Gemini 3 Pro, described as its most intelligent model, to help developers and “vibe coders” bring ideas to life. Built on state-of-the-art reasoning, Gemini 3 Pro delivers stronger results across major AI benchmarks than previous versions and surpasses 2.5 Pro at coding and agentic workflows.

The model excels at multimodal understanding, visual and spatial reasoning, and video reasoning, and offers a 1 million-token context window. It scored 54.2 percentage points on Terminal-Bench 2.0 for tool use via a terminal and tops the WebDev Arena leaderboard with a 1487 Elo, enabling complex instruction following and deep tool use that can translate high-level prompts into interactive apps.

Gemini 3 Pro is available through the Gemini API in Google AI Studio and Vertex AI for enterprise, and it integrates with developer tools such as Google Antigravity, Gemini CLI, Android Studio, Cursor, GitHub, JetBrains, Manus and Cline.

gemini pro, google, gemini api, ai studio, vertex ai, multimodal understanding, visual reasoning, video reasoning, context window, android studio