Google's Gemini App Optimized to Run on Android Phones with Low RAM
Read more
Tecnoblog
tecnoblog.net

Google's Gemini App Optimized to Run on Android Phones with Low RAM

The official Gemini application for Android, launched in early 2024 following the rebranding of the Bard service, has now been optimized to operate on simpler Android devices, even those with only 2 GB of RAM or more.

Previously, users of more basic Android devices were limited to Google Assistant Go, which offered only fundamental commands. However, Google has adjusted the tool so that it functions satisfactorily on devices with reduced RAM or storage capacity.

It is important to clarify that this optimization does not refer to an alternative Gemini application made specifically for entry-level devices; it is the official version of the service that can now operate adequately even with limited memory resources.

Google explained that until recently, hardware restrictions prevented such devices from running the Gemini app, forcing users to rely solely on basic voice commands in the former assistant.

With the app improvement, Gemini began to offer users of these devices features that are already common in more advanced Android phones. Instead of requiring typing extensive phrases, the user can interact by voice, practicing conversations or interviews in over 45 languages through Gemini Live. Additionally, it is possible to use the camera to translate or study notes and forms, or even generate personalized images using Nano Banana, such as a promotional flyer for a bakery.

Although Google has not released a specific list of benefited phone models, the company stated that the revamped version focuses on serving devices that use Android Go, a simplified version of the operating system designed for entry-level devices.

To use the new feature, the device must have at least 2 GB of RAM, as mentioned, and the user must be logged in with their Google Account. This update is global, covering all countries where the artificial intelligence service is accessible.

In the Brazilian context, it is presumed that users of models such as Motorola Moto E14 and Samsung Galaxy A03 will benefit from this functionality. However, certain limitations persist; for example, activation via voice command (such as 'ok, Google') and contextual screen reading may not work on basic Android devices. In these cases, the user has the alternative of starting Gemini by pressing the Home button or the Power button.

Similar stories

Gemini Notebook allows generating interactive websites from notes using simple commands
Read more
tecnoblog.net

Gemini Notebook allows generating interactive websites from notes using simple commands

Gemini Notebook has been updated to enable the creation of interactive websites directly on the artificial intelligence platform, requiring only a simple command. These generated links can be accessed in browsers and shared with colleagues or students, supporting studies in school or university settings and assisting in the development of educational materials.

To begin creating, the user must type a command such as 'Create a website...' in the text box, providing detailed instructions, such as the size of the reading material, which images to include, the type of exercises desired, and other parameters. It is possible to reference books, external links, and the user's own notes, ensuring that the AI consults reliable sources.

This new feature is already available to users over 18 years old and Google Workspace subscribers. Although the company announced that the functionality will reach the mobile application 'soon,' no specific details were provided about this release.

The ability to produce interactive websites is seen as a valuable resource for educational reinforcement. In addition to summaries generated from books and notes defined by the user, it is now feasible to develop quizzes based on a more secure information base.

The request process is quite straightforward: simply specify the necessary sources and notes, ask the platform to create the website, and define the desired style for generating summaries or questions. Upon completion, the website can be adjusted through the prompt box or shared with friends. In a practical example, the resulting website was divided into study phases, summaries about a trip contained in an analyzed book, and a quiz section with linguistic questions, as requested.

Although it is possible to use chatbots like standard Gemini, ChatGPT, or Claude, Gemini Notebook offers specific functionalities to control the queries made by the AI and minimize potential errors. However, it is important to remember that, as it is artificial intelligence, there is a possibility of errors; therefore, it is recommended to verify the information and use as many sources and notes as possible to reduce the occurrence of hallucinations.

Gemini Notebook, which is the current name for the former NotebookLM, is configured as an AI model specialized in assisting with studies. This service allows for the creation of in-depth summaries and detailed analysis of complex topics, facilitating academic routine. Additionally, the platform serves to deepen knowledge in specific professional areas, complementing the understanding of texts and literary works.

Another benefit of the service lies in restricting the sources used, allowing direct use of content saved in Google Books and editing PDF files. Guidelines can be passed directly in the prompt, located on the left side of the screen. Furthermore, it is possible to create notes, summaries, and even content in audio or video format to optimize the learning process.

Google launches Gemini 4 Hydrogen, a new AI model focused on daily tasks
Read more
tecnoblog.net

Google launches Gemini 4 Hydrogen, a new AI model focused on daily tasks

Google announced this Thursday, November 12th, a new artificial intelligence model designed to meet everyday needs. This model, named Gemini 4 Hydrogen, succeeds Gemini 3.8 Flash, promising 2.5 times faster responses and presenting improvements in accuracy.

Currently, Gemini 4 Hydrogen is accessible both in the Gemini app and Google Search, available even to users without a subscription. For developers interested in integrating this model into their own applications and services, the API has a specific pricing structure.

Access to the model can be made through integration tools provided by Google, including Google AI Studio and Android Studio. According to Google, Hydrogen balances speed, reasoning capability, and a lower operational cost compared to other models in its family.

Although it is better suited for text generation and routine tasks, it was observed that the basic version of the model achieved satisfactory results in certain activities. In Artificial Analysis Intelligence, an independent index that evaluates the performance of AI models through various tests, Hydrogen scored 51 points in its default configuration and reached 54 points when using the maximum reasoning level, surpassing the 50 points recorded by Gemini 4 Argon.

Compared to Gemini 3.8 Flash, Gemini 4 Hydrogen demonstrated a reduction greater than five times in the incidence of hallucinations and proved 2.3 times more resistant to prompt injection attacks.

When compared with competing models, such as Anthropic's Claude Sonnet 5.5 and OpenAI's GPT-6.1 Terra, the results indicate certain advantages. However, performance varies depending on the evaluation; in tests requiring system operation and the resolution of scientific and mathematical problems, the Google model lagged behind Claude Sonnet 5.5 in benchmarks such as Terminal-Bench 4.1, Terminal-Bench-Science 0.1, and OSWorld-2.1, a relevant test for agentic AIs that measures interactivity with the computer.

Additionally, the new basic model can be integrated with Gemini 4 Carbon within the same workflow, using the new Orchestrator functionality. This tool operates similarly to the /swarm command, which was introduced in Google Antigravity upon the launch of the frontier model. The strategy is to keep Carbon dedicated to more complex tasks, while Hydrogen handles steps that can be completed with fewer resources. Tests released by Google showed that the combination of these two models achieved success rates close to those obtained by Carbon alone in Terminal-Bench 4.0, resulting in a 24% decrease in token consumption and a 19% decrease in processing time.

Gemini is available on Apple CarPlay, allowing voice conversations with AI while driving
Read more
olhardigital.com.br

Gemini is available on Apple CarPlay, allowing voice conversations with AI while driving

Google has started making the Gemini application directly available on Apple CarPlay. This feature allows iPhone users to interact with the artificial intelligence (AI) by voice while driving, eliminating the need to hold the phone or leave the vehicle's multimedia system interface.

The feature uses Gemini Live, a mode that enables more fluid dialogues with the chatbot. The driver can start a new interaction or resume a previous conversation via an icon displayed on the CarPlay screen. Google stated that it will be possible to pause the dialogue, change the topic, or interrupt the conversation at any time. All conversations are kept synchronized with the Gemini app on the iPhone and recorded in the account history.

The implementation of this feature began this Wednesday, the 11th, and covers iOS users in all languages that support Gemini Live. To take advantage of this novelty, it is essential that the Gemini app is installed and updated on the iPhone, as well as that the user grants microphone usage permission and has initiated at least one conversation with Gemini Live via mobile.

After connecting the iPhone to CarPlay, the Gemini icon becomes visible on the vehicle's interface; simply selecting it starts or continues the voice communication. Prior to the direct integration, Google Assistant was already accessible via CarPlay through the Google Maps app; now, the company offers native access to the chatbot on the car screen.

This addition comes shortly after Google expanded Gemini's presence on Android Auto and in vehicles equipped with the integrated Google system, continuing to introduce personalized assistance features on these platforms.

Popular