Google Launches New Gemini 2.0 Model, Free Deep Research On Android And IOS

Google launched Gemini 2.0, including 2.0 Flash Thinking (experimental), to the web platform, as well as Android and iOS applications. (photo: x @TheAIColony)

JAKARTA - Google is starting to roll out the latest Gemini 2.0 model, including 2.0 Flash Thinking (experimentals), to the web platform, as well as Android and iOS applications. This was discovered after the official announcement on Thursday 13 March.

As usual, Google first launched this latest model on gemini.google.com, before following the mobile app a few days later. As of Saturday afternoon, March 15, the availability of this new model is getting wider on various Gemini accounts.

Gemini 2.0 Flash Thinking And New Changes

All Gemini (free) users can now access 2.0 Flash Thinking (experimental). Google has also started using "experimental" writing in lower and in brackets, including for the 2.0 Pro (experimental) model.

The latest Flash Thinking model is claimed to have better efficiency and speed than the previous version introduced in early February. In addition, Google is now combining app integration features, such as Gmail and YouTube, as well as removing the separate "with apps" model introduced last month.

The file upload feature is also now available to all users, while Gemini Advanced subscribers get a context window of up to 1 million tokens.

Free Deep Research For All Users

The Deep Research feature is now available to all Gemini accounts. This feature uses the Gemini 2.0 Flash Thinking (experimental) model, although Google no longer explicitly mentions the model used.

In the web version, users will find the "Deep Research" button in the prompt bar, or can choose this model via the picker model. However, free users have lower research limits than Advanced subscribers.

Personalization Of AI For Smarter Experiences

Google also introduced a new model called Personalization (experimental), which is available to all users for trial. The model uses 2.0 Flash Thinking (experimental) and can analyze the user search history to provide a more relevant answer.

Some examples of the use of this feature include:

"Where should I go on vacation this summer?" I wanted to start YouTube channel, but needed content ideas. "What should I learn as a hobby or a new job?"

Going forward, Google will expand this feature by using more than just a Search history, including data from YouTube and Google Photos.

Gemini AI Becomes A More Advanced Personal Assistant

Google is increasingly focusing on developing AI assistants who can understand users' needs personally, not just answering general questions.

The Saved Info feature has now been widely launched, while the Advanced capabilities that allow Gemini to remember and refer to previous conversations when answering prompts will soon be available in more than 45 languages.

Gemini 2.0 Model Available For Users

Berikut daftar model terbaru yang tersedia dalam aplikasi Gemini: 2.0 Flash Untuk tugas sehari-hari dengan lebih banyak fitur 2.0 Flash Thinking (experimental) Menggunakan pemalaian tingkat lanjut Deep Research Laporan penyelitian Dalam Personalization (experimental) Bantuan berdasarkan kehistory pencarian 2.0 Pro (experimental) Terbaik untuk tugas yang complex

With this update, Google is further strengthening its position in global AI competition, providing wider access to more sophisticated and personal artificial intelligence technology.