Google has celebrated its annual I/O 2025 eventin which as is happening with the great presentations of other technological ones, AI has had absolute prominence. An artificial intelligence that is no longer simply an assistant to facilitate certain tasks to users, but now goes one step further to become agent and perform various activities on behalf of who communicates the necessary instructions for it.
Google’s AI is, therefore, more advanced and fast. It also offers a higher level of customization to adapt to the needs of users than until now. Above all, for the integration of the company of the company’s multimodal AI models, Gemini, in more products from Mountain View.
In addition, Gemini’s version they will take now is more advanced, since from now Gemini 2.5presented in March, it will be its predetermined model. In addition, Gemini 2.5 carries Deep Think, specialized in jobs that imply code reasoning, and mathematics work.
More creativity thanks to AI
Google has also shown advances in terms of creativity, with I see 3 and image 4tools for creation, respectively, of video and images. I see 3 not only improves the quality of the images that allows you to get your previous version, I see 2, but also now it is capable of generating videos with audio. To create them, it is only necessary to tell you the story of what you want to see and wait a few minutes until you have it ready.
As for image 4, it is characterized by offering high image clarity even in the small details that appear in the images it generates. For example, drops of water or animal hair. Image 4 You can create images with different aspect relationships with a resolution of up to 2K, or more if it is for impressions or presentations. It also stands out in photorealist and abstract styles, and in spelling and typography.
Google has also presented Flowa tool for the creation of films with which it is possible to define all the elements of a history through the use of simple language. Destined to more professional creators, it allows generating scenes, stories and cinematographic clips with the most advanced Deepmind models: I see, image and gemini.
In addition to describing the shots that Flow will have to generate, it will only be necessary to offer details about characters, locations, objects or styles, which can be done from a single point. Flow will be responsible for combining it to generate scenes. For now it is only available for subscribers from Google AI Pro and Ultra in the United States, and will soon reach more countries.
Google ai pro y google ai ultra
Google has also announced two subscription services based on AIthose mentioned Ai pro y ultra. The first offers, for $ 19.99 per month, an AI tools suite, with improvements in the experience of using the Gemini app, and Gemini Advances. It also includes products such as Flow or Notebooklm with special and more relaxed limits and limits than with their free versions.
As for AI Ultra, it will cost no less than $ 249.99 per month. It will give access to the agent mode function, which for now is experimental and will soon reach the desk. With him it is only necessary to say or indicate what is desired, so that Gemini takes care of everything, and will orchestrate the necessary steps to achieve it.
Agent Mode combines advanced functions, such as live web navigation, in -depth research and intelligent integrations, with Google apps. In this way, it allows them to manage complex tasks and several steps from beginning to end, without the subscriber having to do anything apart from supervising it during the process. For now, both subscriptions are available in the United States, and will soon be deployed in more countries.
Google I/O: a more agricultural AI
Google’s AI is no longer limited to giving answers, but is now also able to act on the name of the user, provided that he provides both the instructions of what he wants to do, as well as his permission to do so. Project Marinera system of agents that can perform up to a dozen tasks at the same time. From searches to reservations and purchases, among other things. Of course, for now it will only do so in the United States.
Meanwhile, the Google search engine will have a Way ia (Likewise, for now only in the US), which will allow the browser to do various tasks, such as buying tickets. This will be possible thanks to the integration into him of Project Mariner’s technology.
Project Astra will also arrive at the search engine, which will allow users of a smartphone or a tablet to point in a camera direction, to ask the search engine various issues about what you see through it. The search engine will answer with interactive responses in real time. It will be available soon in AI mode, but only in the United States).
Google will also reinvent, with AI mode, the experience of buying. This mode will allow you to seek online inspiration, test garments in a virtual way and let the agricultural function of the system Buy what you like when your Price is in the range that has been indicated.
Greater customization for AI
To Ia by Google will combine all the knowledge of the world with the context that each user provides. In this way it will work in a more personal way. Gemini Live, for example, is becoming a universal assistant of AI. Little by little it is gaining functions, and soon it can be connected with maps, calendar, tasks and Keep. When it is like this, thanks to the functions of Project Astra that will integrate, you can offer help with a higher customization level.
The IA mode, which we have just mentioned, will make the searches smarter, since thanks to the data it will handle you can offer more adapted recommendations to each user. For example, when looking for information to start preparing a trip to a city, the AI mode will propose ideas for restaurants after searching in the Gmail reservations history of those looking for the information.
Deep Research and Canvas also have more customization, and allow you to merge private files with public data. As for Canvas de Gemini 2.5 Pro, it allows to transform ideas into programming code, applications or creative content. Thus, it allows to create from infographics to podcast -style audio summaries in up to 45 languages.
Canvas is the creative space within the Gemini app, a space to develop and create everything that the user can describe, clearly and easily. Google has also expanded the views created with AI, which will now be available in more than 200 countries, and in more than 40 new languages.
Google Workspace will also incorporate Geminiwhat will do among other things that email has an assistance that will make it smarter, thanks to contextual understanding. In addition, it will allow creating video based on AI from text or slides of a presentation with I see in vids, and having real -time translation with a voice in Meet. Of course, for now only available for Spanish and English, and shortly for French, Italian and Portuguese.
