Google I/O 2025: All the Announcements from the World’s Most Important Artificial Intelligence Event
The tech company highlights Gemini 2.5 Pro, an AI model integrated into Gmail, Chrome, and its search engine. Google I/O 2025 is the event where the company unveils its latest advancements in artificial intelligence. In this edition, Google has focused its attention on this technology, highlighting the launch of the Gemini 2.5 Pro model.
This new model promises improvements in reasoning, multimodal understanding, and coding capabilities. One of its main strengths is its direct integration into products like Gmail, Google Search, and Chrome, allowing users to interact with AI features more smoothly and contextually within these platforms.
“Gemini ranks first in token generation per second—and all while reducing costs. Google is not only leading but has changed the game thanks to our deep investments in infrastructure and talent,” the company stated.
“We’re making intelligence accessible to everyone. For example: a year ago, we processed 9.7 trillion tokens per month. Today, we process 480 trillion—a nearly 50-fold increase!” they added.
Gemini ranks first in token generation per second —and all while reducing costs.
At the I/O 2025 conference, Google highlighted the integration of new AI-based features across its products and services.
One of the new features is the launch of a new “AI Mode” for complex searches, designed to answer longer questions—two to five times longer than usual—and allow sequential follow-up queries. This mode will begin rolling out in the United States.
In the area of video calls, Google introduced Project Beem, the evolution of Project Starline.
This new platform uses artificial intelligence to turn 2D transmissions into real-time 3D experiences. HP will be the first partner to launch compatible devices later this year.
Google Meet will incorporate real-time translation between English and Spanish, with more languages to follow soon, enabling smooth conversations between speakers of different languages.
Google Meet will incorporate real-time translation between English and Spanish.
Another key announcement is Project Astra, an initiative aimed at creating a universal assistant capable of helping with complex tasks such as solving problems, searching for information in emails, manuals, or videos, and even making phone calls.
In the field of content generation, Google highlights Veo 3, a model capable of creating videos with realistic audio, visual effects, and dialogue, and Imagen 4, which improves image and text generation with greater accuracy and detail.
The company also introduced Google Flow, an AI-based cinematic creation tool that allows users to design full narratives—scene by scene—with visual and story coherence.
Gemini Live, another standout feature, will offer camera and screen sharing and will be available for free on Android and iOS devices.
Its capabilities include turning drawings into shopping lists, generating infographics, creating audio in more than 45 languages, and even developing apps from scratch using personal files as references.
Google also showcased Joule, an automated coding agent that can fix errors, create tests, and autonomously update projects.
The company also introduced Google Flow.
Resume about all new AI-Powered Features Across Google Products
- AI Mode for Complex Searches: Designed to answer longer queries (2–5 times longer than usual) and allow sequential follow-ups. Initially launching in the U.S.
- Project Beem: An evolution of Project Starline, transforming 2D video calls into real-time 3D experiences using AI. HP will be the first partner to release compatible devices later this year.
- Real-Time Translation in Google Meet: Supporting English-Spanish translations, with more languages coming soon.
- Project Astra: A universal assistant capable of solving problems, searching through emails, manuals, or videos, and even making calls.
- Veo 3 & Imagen 4: AI models that generate realistic videos, enhance image precision, and improve text accuracy.
- Google Flow: An AI-driven cinematic tool allowing users to design complete narratives, scene by scene.
- Gemini Live: A new feature offering camera and screen sharing, available for free on Android and iOS.
- Google AI Ultra: The best of Google AI in one subscription
Google AI Ultra Subscription Plan.
Besides Google announced a new AI premium plan, providing access to advanced AI features:
- Gemini 2.5 Pro with Deep Think
- Imagen 4 and Veo 3
- Full access to Flow
- YouTube Premium
- 30 TB of cloud storage
This subscription is aimed at digital creators, developers, and AI enthusiasts who want to explore generative models, automate tasks, and leverage cutting-edge AI tools.
Links
- 100 things we announced at I/O – Google Blog
- Sell your Google AI device online – iGotOffer
Ask Search Anything [Video]
Video uploaded by Google on May 20, 2025.
Facebook
Twitter
RSS