Google has as part of its annual developer conference Google I/O 2025 Numerous new products presented-including new functions for the Gemini 2.5 model series, the AI mode in Google Search and a deeper concept of Android XR for a new generation of Smart Glasses with Gemini support. You can find more about this in Google Germany’s blog post “Google I/O 2025: From research into practice” By Sundar Pichai, CEO von Alphabet and Google.
The vision for a universal AI assistant
Google Deepmind develops Gemini 2.5 Pro as a multimodal universal model, which understands plans and experiences through simulation. The goal is a universal, intelligent and context-related AI assistant for all devices through the integration of Project Astra and Project Mariner. So becomes Gemini Live With functions of the research prototype Project Astra expanded, such as language processing, video understanding and memory. Project Mariner Meanwhile, researching how agents can help people with multitasking, especially in the browser. You can find out more about Google’s AI vision in this Blogpost.
Android XR On glasses and headsets
Google also presented the concept of intelligent glasses on the I/O. The goal is direct, context -related support in everyday life based on Android XR and Gemini. Android XR glasses are equipped with camera, microphone, speakers and optional display. The coupling with Gemini hears and understands users: interior interactions and helps to send messages, live translations, appointment agreements or directions. These innovations should enable a new form of AI-based communication. Android XR headsets also become more user-friendly and more powerful by Gemini. You can find more information about this in this Blogpost.
Gemini 2.5 Flash with new functions
Gemini 2.5 Flash encompasses new functions such as native audio output, thought-together (‘Thought summaries’), budgets for thinking models (‘Thinking budgets’), increased security and “deep think”, an expanded mode of thinking for complex mathematics and coding in Gemini 2.5 Pro.
Ki in the search: more than just procurement of information
The search was also equipped with new functions in the course of the I/O. AI fashion in Search offers deeper search results, multimodality and the possibility of immersing them more into a topic through follow -up questions. Within AI Mode, Deep Search can provide even more thorough answers by carrying out hundreds of search queries and summarizing the results as a report.
New AI models and tools for creative
The latest picture position model Image 4 Generates pictures 10 times faster than the previous version. Also comes I see 3 For the first time with native audio generation in the Gemini app. This allows sound effects and background noises and soon even dialogues for the characters in the video.
Flow Is a new AI filmmaker tool from Google based on the AI models Veo, Imagen and Gemini. It is specially tailored to VEO and is characterized by exceptional follow -up of prompt and impressive film maturity.
You can find more details in these blog posts:
OTS original text press release with the exclusive in terms of content of the sender – www.ots.at | SKI