Google Rolls Out Gemini 2.0 AI Models with Free Deep Research
Google has begun a widespread rollout of its latest Gemini 2.0 models, introducing new features to web, Android, and iOS users. Announced on Thursday, these updates include the experimental 2.0 Flash Thinking model, aimed at improving efficiency, reasoning, and personalization. The deployment follows Google’s standard approach, where new models first become available on gemini.google.com before gradually reaching mobile platforms. By Saturday, the availability of these models had significantly expanded across multiple Gemini accounts.
Key Features of Gemini 2.0 Models
Google is enhancing its AI-powered assistant with a suite of new models designed to boost efficiency and usability. The latest updates bring several improvements, including better integration with Google apps, an enhanced file upload system, and new personalization capabilities.
1. 2.0 Flash Thinking (Experimental)
All Gemini (free) users now have access to the 2.0 Flash Thinking (experimental) model, which replaces the previous experimental version launched in February. The updated model boasts improved efficiency and processing speed, allowing users to complete everyday tasks faster.
A notable change in this rollout is how Google presents its experimental models. Instead of capitalizing “Experimental,” the company has now standardized the format as 2.0 Flash Thinking (experimental) and 2.0 Pro (experimental). However, no additional modifications have been announced for the high-end 2.0 Pro model.
One of the most significant updates in this rollout is the integration of Gemini with Google Apps. Previously, Google offered a separate “with apps” model that allowed users to interact with Gemini through Gmail, YouTube, and other Google services. Now, this function has been merged into the standard Gemini experience, ensuring a more streamlined interface.
Additionally, file upload capabilities have been extended to all users, while Advanced subscribers now benefit from a 1-million-token context window, enabling more detailed and complex interactions.
2. Deep Research: Free for All Users
A major highlight of this rollout is the introduction of Deep Research, now accessible to all Gemini accounts. This feature provides detailed, in-depth research reports and advanced analytical insights. While it is powered by the 2.0 Flash Thinking (experimental) model, Google no longer explicitly states which model is being used for Deep Research.
Users can access Deep Research through a new button in the prompt bar on the web, making it easier to generate comprehensive reports. Alternatively, they can continue using the model picker to select the desired AI capabilities. However, free users face more restricted research limits compared to Advanced subscribers.
3. Personalization (Experimental)
Another significant update is the introduction of Personalization (experimental), a new feature available to all Gemini users. This AI-driven capability leverages a user’s Google Search history to enhance responses based on past queries.
For example, if a user searches for travel suggestions, the model will tailor recommendations based on their previous preferences. Similarly, if someone is brainstorming for a YouTube channel, Gemini can generate customized content ideas based on past interactions.
Here are some example prompts that showcase this new personalization feature:
-
Where should I go on vacation this summer?
-
I want to start a YouTube channel but need content ideas.
-
What would you suggest I learn as a new hobby or job?
While personalization currently relies solely on Google Search history, Google has hinted at future expansions that will include data from YouTube and Google Photos. The broader goal is to create an AI assistant that not only answers general questions but also understands individual user preferences, making interactions more intuitive and personalized.
4. Saved Info and Expanded Recall Capabilities
As part of its broader AI initiative, Google has fully rolled out Saved Info, which allows Gemini to retain key details for reference in future conversations. Additionally, Advanced subscribers will soon have access to a powerful recall feature that lets Gemini remember and reference past chat interactions when generating responses. This capability will be available in over 45 languages, significantly expanding Gemini’s usability for global users.
The Latest Gemini 2.0 Models Available
With this rollout, Google has made five key AI models available in the Gemini app:
-
2.0 Flash – Designed for everyday tasks with enhanced processing capabilities.
-
2.0 Flash Thinking (experimental) – Features advanced reasoning for more complex interactions.
-
Deep Research – Provides in-depth analytical reports for detailed research projects.
-
Personalization (experimental) – Offers tailored responses based on Google Search history.
-
2.0 Pro (experimental) – Optimized for complex problem-solving and high-end AI tasks.
A Step Toward Smarter AI
This latest update reflects Google’s ongoing efforts to improve the capabilities of its Gemini AI assistant. By making Deep Research freely available, expanding file uploads, and introducing personalized AI experiences, the company is positioning Gemini as a powerful tool for both casual users and professionals.
The inclusion of advanced memory and recall capabilities also signals Google’s long-term ambition: creating an AI assistant that not only provides accurate answers but also remembers user preferences to deliver context-aware suggestions.
As Google continues to refine its AI models, these enhancements will likely play a crucial role in shaping the next generation of AI-powered assistants, making digital interactions more personalized, efficient, and insightful than ever before.
Also Read: Google Unveils Gemma 3: AI Model Optimized for Single GPUs
