After announcing the launch of Gemini 2.0 at the end of last year, Google earlier announced the launch of the further upgraded Gemini 2.5, and initially provided the fully functional Gemini 2.5 Pro Experimental version, which is touted as having considerable capabilities in programming, mathematical reasoning, and scientific analysis.
Koray Kavukcuoglu, Chief Technology Officer of Google DeepMind, pointed out that Gemini 2.5 has significantly improved its inference capabilities with a significantly enhanced basic model and improved post-training, and can handle complex problems and serve as a more powerful artificial intelligence agent.
The first version launched this time is the Gemini 2.5 Pro experimental version. Unlike the lighter Flash Thinking version launched when Gemini 2.0 was first released, it is obvious that it hopes to compete with other competitors in the market, especially artificial intelligence models such as DeepSeek R1 that have recently emerged in China. It is also expected that more derivative versions will be launched in the future.
Currently, the experimental version of Gemini 2.5 Pro will be available through Google AI Studio. Users who pay for the Gemini Advanced service can also use the Gemini 2.5 Pro experimental version model directly in the Gemini App.
