Google launches Gemini 2.5, its most intelligent AI model: CEO Sundar Pichai says ‘start of a new era of thinking models..’

has launched Gemini 2.5, labelling it the company's “most intelligent AI model” to date. The initial release is an experimental version of Gemini 2.5 Pro, which Google reports has achieved state-of-the-art results across various benchmarks and secured the top spot on the LMArena leaderboard, a platform measuring human preference for AI model responses.
Company officials describe the new Gemini 2.5 models as “thinking models,” emphasising their enhanced capability for reasoning through complex problems before generating a response. This approach, Google claims, leads to improved performance and accuracy.
“Our first 2.5 model, Gemini 2.5 Pro Experimental is a state-of-the-art thinking model, leading in a wide range of benchmarks – with impressive improvements in enhanced reasoning and coding and now #1 on @lmarena_ai by a significant margin. With a model this intelligent, we wanted to get it to people as quickly as possible,” Google CEO Sundar Pichai said in a post on X.
“Find it on Google AI Studio and in the @geminiapp for Gemini Advanced users now – and in Vertex in the coming weeks. This is the start of a new era of thinking models – and we can’t wait to see where things go from here,” he added.


This development builds on Google's previous research into techniques like reinforcement learning and chain-of-thought prompting, as well as the recent introduction of Gemini 2.0 Flash Thinking. The company stated that Gemini 2.5 achieves its performance gains through a combination of a significantly upgraded base model and improved post-training processes. Google plans to integrate these "thinking" capabilities directly into all future models.
The experimental Gemini 2.5 Pro is positioned as Google's leading model for complex tasks. The company highlighted significant advancements in coding capabilities compared to its predecessor, Gemini 2.0.
Gemini 2.5 Pro retains features from previous Gemini generations, including native multimodality (handling text, audio, images, video, and code) and a large context window, starting at 1 million tokens with plans to expand to 2 million soon.
Developers can access the experimental Gemini 2.5 Pro immediately via Google AI Studio, while Gemini Advanced subscribers can select it within the Gemini app.