March 25, 2025 — Gemini 2.5 is here to assist you with solving considerably more complex problems. It is a thinking model with advanced reasoning and coding capabilities. The new model has been launched as an experimental version of 2.5 pro with ‘state-of-the-art‘ outcomes. Google has addressed Gemini 2.5 as their ‘most intelligent AI model‘ to date, and it is now available for Gemini Advanced users in Gemini App and the Google AI Studio.
With a focus on greater performance and increased accuracy, Google has designed 2.5 thinking models with improved reasoning abilities. The previous models relied on chain-of-thought prompting and reinforcement learning. However, the formulation of thinking models, including Gemini 2.0 Flash Thinking, has offered new possibilities with reasoning power and resolving increasingly complicated queries. Gemini 2.5 will soon be available on Vertex AI. Google has not announced its pricing yet.
Gemini 2.5 includes native multimodality with the capability to process larger datasets while offering resolutions in multiple formats, including text, image, audio, video, and even entirely in codes. With the launch, the new model shipped one million token context windows, which is the highest for any Gemini model till now. Soon, the ratio will be increased to two million tokens.
With increased reasoning capacity, the new model has achieved remarkable breakthroughs. The astonishing results have led Gemini 2.5 to make a remarkable debut on LMArena at rank #1 with the current score of 1443.
Sunder Pichai, Google CEO, has expressed on X, “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.”
He further adds, “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.”
What’s New in Gemini 2.5?
Enhanced Reasoning:
Gemini 2.5 includes next-level reasoning power, setting benchmarks in various fields like science and mathematics. The model outperformed GPQA and AIME 2025, renowned math and science benchmarks. Alongside that, it has become the first experimental model to achieve 18.8% in the knowledge and reasoning benchmark, Humanity’s Last Exam, leaving other models like OpenAI o3-mini behind.
Advanced Coding:
Gemini 2.5 brings advanced coding and programming capabilities even better than the previous 2.0 version. The 2.5 Pro version can transform and edit codes, formulating intuitive applications. In the custom agent setup, the thinking model attained 63.8% in the Agentic coding benchmark, SWE- Bench Verified.
Why Does it Matter?
AI advancements have changed the way we contemplate technology today. Alongside that, it has also streamlined how we solve complex tasks with efficiency. AI integrations are everywhere, from our personal to professional lives, posing significant prospects for the upcoming years. Hence, it is necessary for us to stay aware of each progression in this field.
YourTechDiet is determined to bring the latest in tech to you at the earliest. Check out our in-depth content repository to stay informed in the digital era!
Also Read: Virtual assistants vs. Chatbots: What’s the Difference