The Google Gemini app gained new AI features like Deep Research and the experimental 2.0 Flash Thinking model.
In December 2024, the search engine company released Gemini 2.0 Flash, which supports multimodal inputs and outputs.
READ: Google Gemini 2.0 prepares for a future of AI agents
It can generate native images with text and produce customizable text-to-speech multilingual audio.
Recently, Google improved its reasoning capabilities and provided a 1-million token context window for Gemiini Advanced users.
Notes, Tasks, Photos, Calendar, and other apps that link to Gemini will roll out on 2.0 Flash Thinking Experimental.
As a result, it can tackle complicated tasks that typically require multiple apps to accomplish. For example, Google says it can handle the following query:
“Look up an easy cookie recipe on YouTube, add the ingredients to my shopping list, and find me grocery stores that are still open nearby.”
Google Gemini 2.0 Flash also improves Deep Research, an AI model that studies, collects, and compiles data into comprehensive reports.
Lastly, the improved Google Gemini app lets you create custom AI agents called Gems or choose premade ones.
Open the app’s Gems Manager on the desktop, write instructions, give it a name, and then chat with it whenever. Also, you may upload files into your custom Gem as its reference.
Soon, Google Gemini will integrate with Google Photos so you can ask for help with your pictures.
For example, you may ask the AI model to show pictures from your recent trip. Then, it can create a travel plan based on the places in these images.
