Google has introduced Gemini 2.5, the latest iteration of its artificial intelligence model, which the company describes as its most intelligent yet. The new model, particularly the Gemini 2.5 Pro Experimental version, has demonstrated significant improvements in reasoning, coding, and mathematical capabilities, according to Google.
“Gemini 2.5 Pro is available now in Google AI Studio and in the Gemini app for Gemini Advanced users, and will be coming to Vertex AI soon. We’ll also introduce pricing in the coming weeks, enabling people to use 2.5 Pro with higher rate limits for scaled production use,” Google said in a statement.
The Gemini 2.5 models are being positioned as “thinking models,” a term Google uses to describe AI systems designed to analyze information, process context, and engage in complex reasoning before generating responses. This approach, the company says, enhances accuracy and overall performance, setting a new standard for AI-driven decision-making.
“For a long time, we’ve explored ways to make AI smarter and more capable of reasoning through techniques like reinforcement learning and chain-of-thought prompting,” Google added. “With Gemini 2.5, we’ve achieved a new level of performance by combining a significantly enhanced base model with improved post-training.”
Benchmark Leader
Google claims that the Gemini 2.5 Pro Experimental model has topped the LMArena leaderboard, a benchmark that measures human preference in AI-generated outputs. Additionally, it has outperformed prior models in tests assessing math, science, and coding proficiency, including leading results on datasets such as GPQA and AIME 2025.
In a notable leap from its predecessor, Gemini 2.0, the new model has reportedly shown a major boost in coding abilities, particularly in generating and editing software applications. It achieved a 63.8% score on SWE-Bench Verified, a key benchmark for assessing agentic coding capabilities.
Beyond code generation, the model’s ability to process vast amounts of information is another highlight. Google says Gemini 2.5 Pro currently supports a context window of one million tokens, with plans to expand to two million, allowing it to comprehend extensive datasets across multiple formats, including text, audio, images, video, and even entire code repositories.
Deployment and Accessibility
Gemini 2.5 Pro is now available in Google AI Studio and for Gemini Advanced users through the Gemini app. Google also announced plans to roll out the model on Vertex AI, its cloud-based AI service, in the coming weeks. The company will introduce pricing models soon, allowing users to scale their use of the model for enterprise applications.
As AI models become more advanced, the emphasis on reasoning, contextual awareness, and multimodal capabilities continues to define the competitive landscape. With Gemini 2.5, Google is positioning itself at the forefront of this evolution, as it seeks to build AI systems that are not only more powerful but also more adaptable to real-world complexities.
