Google has rolled out Gemini 3, its newest artificial intelligence model, in a move that signals how aggressively the company is pushing to keep pace with OpenAI’s rapid advances in generative AI.
Sundar Pichai, CEO of Alphabet, said the upgraded model delivers stronger answers to complex queries and reduces the amount of prompting users need before getting accurate results. The rollout begins with select subscribers and will expand widely in the coming weeks. Gemini 3 is being baked directly into the Gemini app, Google’s AI search features such as AI Mode and AI Overviews, and a range of enterprise-facing tools.
This launch arrives only eight months after Gemini 2.5 and less than a year after Gemini 2.0—an unusually fast development cycle that underlines how quickly the major AI companies are iterating in response to one another. OpenAI introduced GPT-5 in August and updated the model again last week, adding a version the company said feels “warmer” and “more intelligent,” as well as another designed to work faster on simple tasks and stay focused on complex ones.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
Pichai wrote that AI has evolved in two years from reading text and images to now understanding complex intent, tone, and context.
“Starting today, we’re shipping Gemini at the scale of Google,” he said, adding that the new model has been engineered to understand nuance at a deeper level and anticipate the meaning behind users’ requests.
Google said its ecosystem continues to expand at a rapid clip. The Gemini app now records 650 million monthly active users, while AI Overviews sees 2 billion monthly users. OpenAI reported in August that ChatGPT has 700 million weekly users, keeping the two companies neck-and-neck in overall reach.
Demis Hassabis, CEO of Google DeepMind, said Gemini 3 will aim for more direct and substantial answers. He added that the system is designed to avoid empty or flattering responses, instead prioritizing insight and clarity—an answer to long-running complaints that many chatbots tend to be overly agreeable.
The tech giants building these systems have been pouring unprecedented sums into the hardware and data-center capacity required to sustain them. Alphabet, Meta, Microsoft, and Amazon each raised their capital expenditure forecasts during their latest earnings calls and together expect to spend more than $380 billion this year.
Google is pairing the new model with an expanded developer ecosystem. It introduced “Google Antigravity,” a platform that lets developers code at a higher task-driven level. The company described Gemini 3 as its strongest model yet for vibe coding, a term referring to tools that let developers generate software through conversational prompts.
The company also showcased “generative interfaces,” a new experience where the model can produce answers that resemble a digital magazine, mixing images, stylized text, and structured layouts. In one demonstration, the model was asked to explain the Van Gogh Gallery with added life context, and the response produced a visually rich, mixed-media overview of each piece.
Paid subscribers will be the first to try these enhanced capabilities within AI Mode. Gemini 3 can analyze a user’s question and build tailored outputs such as interactive tables, grids, or custom tools. Google said the model can assemble a loan calculator on the fly or build simulations that help explain complicated physics problems.
Developers will gain access through the Gemini API, while companies can integrate the system through Vertex AI, Google’s cloud platform for deploying AI models. For corporate users, the company said Gemini 3 can be deployed for employee onboarding, procurement tasks, and industrial operations, including analyzing videos and images from factory floors with higher accuracy.
The unveiling positions Google for another round of head-to-head competition with OpenAI, with both companies racing to dominate consumer apps, enterprise tools, and developer ecosystems. The intensity has been rising throughout the year, and Gemini 3 gives Google another major step as the rivalry widens across search, productivity software, creative tools, and coding environments.
This latest launch signals that the speed of innovation in generative AI is unlikely to slow down soon. Each release from Google or OpenAI is now prompting near-immediate responses from the other, setting up a rivalry that continues to deepen across products, platforms, and user bases.