Google’s annual I/O event this week put AI and XR front and center, showcasing the company’s accelerating work in these areas. Google executives demonstrated how its Gemini AI models have seen over a 50x increase in monthly token usage in the past year, with a significant inflection point being the release of Gemini 2.5 in March 2025. This development has supercharged Google’s AI efforts, particularly in the agentic era of AI.
The company also reiterated its commitment to Android XR, releasing the second developer preview and expanding Gemini’s reach into various platforms, including Google TV, Android Auto, and WearOS. Android XR, being the first OS from Google built from the ground up during the Gemini era, integrates AI from the outset to enhance user experiences.
To accelerate on-device AI adoption, Google announced improvements to LiteRT, its runtime for local AI models, and introduced the AI Edge Portal for developers to test and benchmark on-device models. These advancements are crucial for low-latency and secure user experiences.
Google also made significant strides in XR hardware through partnerships. The company reaffirmed its commitment to Samsung and Project Moohan, expected to ship this year, and announced collaborations with Warby Parker and Gentle Monster to develop smart glasses. Additionally, Google unveiled Project Aura, a headset developed in partnership with Xreal, aiming to capture a middle ground between immersive headsets and lightweight smart glasses.
Google Beam, formerly Project Starline, received an update with near-real-time translated conversations, enhancing the conferencing experience. This feature, available in Google Meet, captures a person’s tone, expressions, and accents while translating their speech.
Analysts believe Google understands the intersection of AI and XR and has the capabilities to leverage this understanding. The company’s approach to XR is now more aligned with its overall efforts, particularly with the growth of AI and Gemini’s capabilities. AR glasses are seen as a natural way to experience AI, offering a more immersive and interactive experience than smartphones.