Google rolling out Gemini 2.0 Flash model to Gemini app, web: What's new

Google said Gemini 2.0 Flash can provide help for everyday tasks like brainstorming, learning or writing

Google has started rolling out a stable version of its new Gemini 2.0 Flash AI model to the Gemini app on smartphones and Gemini on the web. The first stable version of a second-generation Gemini model has now started appearing in the model picker of Gemini AI for some users. Google describes it as a model of choice "For everyday tasks."

 

After launching Gemini 2.0 in December last year, Google made an experimental Gemini 2.0 Flash model available to users for testing. However, upon choosing the experimental model it clearly stated that it is currently under testing and it "might not work as expected."

 

Gemini 2.0 Flash: What is new

 

In an update to the Gemini blog, Google said that the stable version of the Gemini 2.0 Flash delivers fast responses and stronger performance across a number of key benchmarks, providing everyday help with tasks like brainstorming, learning or writing. With the new model, Gemini Advanced subscribers will have access to a 1 million token context window for up to 1,500 pages of file uploads. Paid-tier subscribers can now also use the new Gemini 2.0 Flash in custom Gems.

At the time of writing, the Gemini 2.0 Flash model has started to appear on Gemini's web version. However, the smartphone app on both Android and iOS shows the experimental version of the model. Wider roll-out on the Gemini mobile app is expected soon.

 

Imagen 3 in Gemini

 

Besides the Gemini 2.0 Flash model integration into Gemini, Google has also updated Gemini AI's image generation capabilities with Imagen 3. Previewed at the company's annual developers conference, Google I/O, last year, Imagen 3 is the latest imaging model by Google. The company said that it brings a higher degree of photorealism, better instruction following, and fewer distracting artefacts.