Tech Desk 
Mar 27, 2025
Google has introduced Gemini 2.5, its latest and most powerful AI model, designed to improve reasoning and problem-solving. The company also launched Gemini 2.5 Pro Experimental, a multimodal AI model that Google calls its smartest yet. This advanced AI can "pause to think" before responding, making it more accurate and efficient.
Gemini 2.5 Pro is now available on Google AI Studio for developers and in the Gemini app for users subscribed to the $20-a-month Gemini Advanced plan.
Gemini 2.5 Pro to offer better performance in AI tests
Google claims that Gemini 2.5 Pro outperforms previous models and many top competitors. In a coding test called Aider Polyglot, it scored 68.6%, beating OpenAI, Anthropic, and DeepSeek models.
In another test, SWE-bench Verified, which measures software development skills, it scored 63.8%, performing better than OpenAI’s o3-mini but slightly behind Anthropic’s Claude 3.7 Sonnet.
The model also did well in a multimodal exam called Humanity’s Last Exam, scoring 18.8% higher than most AI models in answering math, science, and humanities questions.
One of Gemini 2.5 Pro’s biggest upgrades is its 1 million token context window. This allows it to process 750,000 words at once, longer than the entire “Lord of the Rings” series. Google also plans to expand this to 2 million tokens soon.
While the company has not yet shared API pricing for Gemini 2.5 Pro, more details are expected in the coming weeks. With its improved reasoning and expanded context, Gemini 2.5 is set to redefine AI-powered problem-solving.
