Google has begun the rollout of its latest Gemini 2.0 Flash artificial intelligence (AI) model for the Android version of its chatbot. The initial model from the Gemini 2.0 series was introduced on December 12, coinciding with its release on the web. However, mobile applications did not gain immediate access. Users can now expect to see this model being integrated into their apps, alongside a new feature that allows them to select different AI models.
Gemini 2.0 Flash AI Model Now Available on Android
This advancement in the Gemini AI models comes nine months following the debut of the Gemini 1.5 series. Google has asserted that this new generation enhances overall capabilities, including native image support and generation functionalities. The Flash variant, characterized as the smallest and fastest model in the series, is currently accessible in an experimental preview stage.
Model switcher feature in the Gemini app
Users on the Google app for Android version 15.50 beta will soon notice two significant changes within the Gemini app. The first is that the model information at the top of the screen has become interactive. A downward arrow has been added between ‘Gemini’ and ‘1.5 Flash’ for free users, serving as the model switcher. This feature has been verified by staff members at Gadgets 360.
The second notable update includes the introduction of the Gemini 2.0 Flash experimental model. Tapping the model switcher reveals a bottom sheet that displays the available AI models. Free users will find only the 1.5 Flash and 2.0 Flash options, while subscribers of Gemini Advanced will also have access to the 1.5 Pro model.
It’s important to note that Google has mentioned the Gemini 2.0 Flash is currently available as an early preview, which may not function optimally. Some features of Gemini might also be incompatible with this AI model until a stable version is launched.
Upon its release, Google asserted that Gemini 2.0 Flash surpassed the 1.5 Pro model in various benchmark tests conducted during internal evaluations. These tests included assessments such as the Massive Multitask Language Understanding (MMLU), Natural2Code, MATH, and the Graduate-Level Google-Proof Q&A (GPQA).
