Google’s Gemini app launches 2.0 Flash AI model

Google has announced that its Gemini app is now utilizing the Gemini 2.0 Flash AI model, promising faster responses and improved performance for various tasks such as brainstorming, learning, and writing.

Google announced today that its Gemini app is now using the Gemini 2.0 Flash AI model, promising faster responses and better performance for a variety of tasks, including brainstorming, learning, and writing. The rollout is happening on both web and mobile, and all users will have access to the updated model.

Google’s Gemini app gets 2.0 Flash AI model

When Google introduced Gemini 2.0 in December, it said that it was working to quickly integrate this model into its products. An experimental version was made available to users before, but now Gemini 2.0 Flash is the default for the app.

In addition to the performance improvements, Google said that it has improved Gemini’s image generation capabilities with the latest version of Imagen 3, the AI text-to-image generator. This new model is designed to generate images with richer details and textures, while following user instructions more precisely.

Video: Google

Gemini Advanced users get more features

As a reminder, Gemini Advanced users get a 1M token context window and can upload files of up to 1,500 pages, plus they get priority access to features like Deep Research and Gems. The Gemini 1.5 Flash and 1.5 Pro models will still be available for users for the next few weeks for ongoing conversations.

In a related update, some users noticed that references to the Gemini 2.0 Pro Experimental model were removed from the changelogs. A Google spokesperson clarified that an outdated note was published in error. The Gemini 2.0 Pro Experimental model, which is available to Gemini Advanced users, is Google’s top AI model, reportedly offering better factuality and performance in coding and mathematics tasks.

This experimental model is designed to help users with advanced tasks, such as generating specific programming code or solving advanced math problems. However, Google notes that it is in an early preview stage and may behave unexpectedly. It will not have access to real-time information and is not compatible with some features in the app.

Main image credit: Google