Google has launched Gemini 3, its latest foundation model. It is now available through the Gemini app and AI search. This update arrives just seven months after Gemini 2.5, putting Google at the forefront of the competition for the most capable AI system today.
The timing is significant, coming just days after OpenAI released GPT 5.1 and two months after Anthropic introduced Sonnet 4.5. This highlights the quick pace of advancements in model development.
More advanced AI Model Gemini 3 launched
A more research-focused version called Gemini 3 Deepthink is also coming soon. It will be available to Google AI Ultra subscribers in the next few weeks after passing additional safety checks.
“With Gemini 3, we’re seeing this massive jump in reasoning,” said Tulsee Doshi, Google’s head of product for the Gemini model. “It’s responding with a level of depth and nuance that we haven’t seen before.”
Early benchmark scores back up these statements. With a score of 37.4, Gemini 3 now ranks at the top of Humanity’s Last Exam benchmark, which measures high-level reasoning and general knowledge. The previous record was held by GPT 5 Pro with a score of 31.64. The model has also reached the number one position on LMArena, a leaderboard based on user satisfaction.
650 million monthly active users already use
Google reports that the Gemini app currently has over 650 million monthly active users, and 13 million developers are integrating the model into their workflows. These numbers show how important Gemini has already become to Google’s larger AI ecosystem.
Alongside the main model release, Google introduced Antigravity, a new coding interface powered by Gemini. It offers multi-pane coding to developers by combining a chat-style prompt window with a terminal and a browser view. This setup allows the coding agent to write, test, and refine applications in real-time.
Read also: Shubhanshu Shukla shares latest update about Gaganyaan Mission launch