Instead of holding a big launch event, the company quietly mentioned the model in an update to its Gemini chatbot app.
The Gemini 2.0 Pro Experimental is the successor to the Gemini 1.5 Pro, which was announced in February last year. The arrival of this AI model comes amid the tech industry's attention to DeepSeek, an AI startup from China.
DeepSeek’s latest models, which are widely available for download and use, are performing on par with or even surpassing many products from other major tech companies, prompting Silicon Valley and the U.S. government to rethink the AI race.
According to Google, Gemini 2.0 Pro Experimental is now the flagship AI model in the Gemini ecosystem and will offer higher accuracy, as well as more powerful processing of programming and mathematical tasks.
In the update description, Google writes: "Whether you're tackling complex programming challenges, like creating a complete program from scratch, or solving mathematical problems like developing statistical models or quantum algorithms, Gemini 2.0 Pro Experimental helps you perform these tasks with greater accuracy and efficiency."
The model is available to Gemini Advanced users since late January and is included in the Google One AI Premium paid plan as well as Gemini add-ons for Google Workspace.
Google emphasizes that Gemini 2.0 Pro Experimental is still in early testing and may exhibit unexpected behavior or bugs. Unlike other Gemini models on the app, this version cannot access real-time information and does not support some app features.
“We believe in rapid innovation and want to bring the best of Gemini technology to the world. Gemini Advanced users will have priority access to our latest AI innovations. Your feedback helps us improve our models over time and optimize how we deploy AI more broadly.”
Along with the launch of Gemini 2.0 Pro Experimental, Google is also bringing the Gemini 2.0 Flash version - which was announced in December - to the Gemini app for all users. This will be the default model of the app going forward.