Google launches Gemini 3.5 Flash and new virtual agent to intensify competition in the technology market

Alphabet, the parent company of Google, announced the integration of new artificial intelligence tools aimed at continuous automation and daily productivity. The presentation event highlighted the launch of the Gemini 3.5 Flash model and a new virtual agent. The company’s strategy focuses on reducing operational costs and increasing processing speed. Desenvolvedores and large corporations represent the main target audience of the ecosystem update.

The movement occurs at a time of high pressure in the global technology market. Investidores demand practical results after investing billions of dollars in research and server infrastructure. The new product line aims to consolidate complex functionality directly into everyday applications. The integration seeks to strengthen the brand’s presence in users’ digital routine and optimize corporate tasks on a large scale.

We asked our agents to build a working operating system from scratch using @Antigravity 2.0 and Gemini 3.5 Flash.

It took:
⏱️ 12 hours
🤖 93 parallel sub-agents
🔄 15k+ model requests
🧠 2.6B tokens processed
💸 Less than $1K in API credits

To build a functioning OS from… pic.twitter.com/DRwflfF2eK

— Google (@Google) May 19, 2026

Gemini 3.5 Flash takes over the center of Google services

The new model serves as the default engine for most of the company’s applications and operating systems worldwide. Sundar Pichai, the company’s chief executive, detailed the architecture’s efficiency improvements during the conference. The tool processes information up to four times faster compared to previous systems. Reducing latency allows for more natural interactions and almost instantaneous responses. Waiting times have dropped drastically.

Software developers receive direct access to the updated system through dedicated programming interfaces. The modified framework makes it easier to write complex code and build dynamic conversational environments for customer service. The operating cost presents a significant drop compared to the past generation of artificial intelligence. The amount charged for using the platform reaches around a third of the price charged by direct competitors in the current market. The change positions small and medium-sized companies at the forefront of technological access, allowing the creation of robust solutions without the need for million-dollar budgets.

The platform’s security has undergone rigorous adjustments to avoid unnecessary blockages during everyday use. Old Filtros used heavy algorithms that interrupted processes without clear justification. The recent version understands the context of questions more accurately. The system reduces barriers to harmless content and eliminates erroneous automatic refusals. The version aimed at high-level users is undergoing internal testing and is scheduled for release in June.

Omni System introduces physics simulation and advanced processing

Understanding the real world takes on a new dimension with the sale of models focused on physics and three-dimensional space. The software analyzes videos, audios and images simultaneously. The technology predicts the behavior of objects and people in varied scenarios. The system replaces old tools and expands creative possibilities for users involved in the production of audiovisual content. Multimodal processing allows the machine to understand the relationship between sound and movement natively.

Profissionais video features are built directly into the platform to modify recorded clips with cinematic precision. Using simple text commands adds complex visual elements, changes scene lighting, or modifies character trajectories in real time. One of the main innovations solves the chronic problem of visual consistency between frames generated by artificial intelligence. Characters’ physical characteristics, such as clothing and facial expressions, remain intact even after multiple edits to the original video file. Essa stability reduces post-production time in animation studios and advertising agencies.

Leia Tambem

United States classifies PCC and Red Command as terrorist groups with blockade of assets

United States classifies PCC and Red Command as terrorist organizations in decision from June 5th

Luxury meeting financed by Daniel Vorcaro in New York becomes the target of Federal Police investigation

The integration of conversational features is now available to paid plan subscribers. The application of technology is not limited to the creation of videos or digital texts. Pesquisadores uses databases to train robots in simulated environments before deployment in the physical world. Support for this file format will be gradually expanded over the next few weeks.

Autonomous virtual Agente performs tasks in the cloud

The level of personal automation reaches a new level with the arrival of the cloud-based assistant, called Spark. The program operates continuously on virtual machines dedicated to users. The system works independently and allows people to delegate long tasks while closing the application. Execution occurs in the background and eliminates the need to wait for complex processing to finish.

The assistant analyzes data from several connected applications simultaneously and makes logical decisions based on the user’s historical preferences. Planning complex international trips, organizing overflowing inboxes, and writing financial reports all occur without constant supervision. Integration with the company’s corporate platform expands the reach of automated actions in the near future, connecting spreadsheets, documents and calendars. The number of clicks and manual commands required in the digital routine decreases considerably, freeing up time for strategic activities.

The testing phase covers a restricted group of participants before commercial launch. The company has started initial bug fixes and stability calibration. The use of autonomous agents represents a fundamental change in the sector’s business model. The technology transforms the traditional user interface into an environment where artificial intelligence anticipates demands.

Disputa for leadership drives updates in the sector

The technology market is facing a phase of intense competition and consolidation among the largest global corporations headquartered in Vale of Silício. Empresas major rivals, such as OpenAI and Anthropic, maintain an aggressive pace of software releases, driven by record financial valuations and venture capital contributions. Google uses the strategic advantage of having the world’s leading server infrastructure to quickly deploy new functions on a global scale. Retaining billions of users in its ecosystem guarantees the immediate application of tools on smartphones, browsers and operating systems.

Engagement numbers indicate accelerated adoption of new technologies by consumers in several regions. The flagship app has surpassed 9 million recent active users. The volume of data processed monthly reaches 9.7 billion requests. The scale of the operation requires massive investments in equipment and data center cooling.

• The default model responds readily to text commands and daily information collection.
• The system provides accurate simulations for content creators and robotics researchers.
• The Spark agent focuses on corporate productivity and the automation of individual processes.
• The professional version remains in the adjustment phase and is expected to be released at the end of the semester.
• Prices vary depending on the subscription level chosen by consumers in the premium package.

The renewal of the user interface accompanies the change in processing engines to facilitate intuitive use. The corporate ecosystem receives native integrations and drives the adoption of tools by large companies and educational institutions. The evolution of the system demonstrates a scenario where technology operates constantly behind the digital scenes. Execution of complex commands occurs with minimal direct interference from the end user.