Google's Gemini 3 Pro Takes Lead in AI Benchmarks by 3 Points

Google has officially launched Gemini 3, its latest and most powerful artificial intelligence model, which the company states sets new industry standards for multimodal understanding and agentic capabilities. The announcement, made on November 18, 2025, positions Gemini 3 Pro as the new leader in AI, surpassing competitors by an aggregate of three points in independent analyses. Alphabet CEO Sundar Pichai highlighted that the new model allows users to achieve desired results with "less prompting."

Gemini 3 brings significant advancements in reasoning and multimodal understanding. Its "Deep Think" mode achieved a 41.0% score on Humanity's Last Exam without tools and an impressive 93.8% on GPQA Diamond, showcasing its enhanced problem-solving abilities. Furthermore, the model scored 54.2% on Terminal-Bench 2.0, a benchmark for tool-use in operating a computer via terminal, indicating robust coding agent capabilities.

The new AI model is being integrated across Google's ecosystem, including the Gemini app, AI Mode in Search, AI Overviews, and its enterprise products. Developers can immediately access Gemini 3 through Google AI Studio, Vertex AI, and Gemini CLI. This widespread rollout on day one signifies Google's commitment to delivering its latest innovations at scale.

Accompanying Gemini 3 is the introduction of "Google Antigravity," a new agentic development platform designed to enable developers to code at a higher, task-oriented level. This platform leverages Gemini 3's advanced reasoning and tool-use capabilities, allowing agents to autonomously plan and execute complex software tasks. Josh Woodward, Vice President of Google Labs and Gemini, described Gemini 3 as the company's "best vibe coding model ever."

The launch intensifies the competition within the AI sector, particularly with OpenAI, and comes amidst ongoing discussions about the "AI bubble." Analysts from Artificial Analysis have confirmed Gemini 3 Pro as the new leader in AI, noting its aggregate score three points ahead of GPT-5. This significant leap in capabilities suggests that the rapid pace of AI innovation continues unabated.