Google Shifts Pixel AI to Local Data Processing

Google continues to expand artificial intelligence capabilities on Pixel smartphones, focusing on local data processing. The company introduced the new Gemma 4 E2B model, specifically adapted for the Tensor chip used in Pixel devices. This technology allows applications to perform tasks directly on the smartphone without sending requests to the cloud, accelerating performance and ensuring functionality even without an internet connection.

At the Google I/O India event, practical use cases for local AI were demonstrated. For example, Pixel 10 users will be able to plan trips, receive recipe recommendations, control smart home devices, and use Mobile Actions voice commands to perform tasks on their smartphones without internet access. These features are particularly useful in areas with unstable connectivity or when operating offline.

Additionally, Google emphasizes the multimodal capabilities of AI. Pixel 10 smartphones will be able to recognize objects in photos, such as landmarks or plants, and transcribe lectures and voice notes without an internet connection. In the corporate sector, the technology can be used to create offline maps of retail spaces or identify faulty car parts from photos.

The development of local AI models is part of Google’s strategy to expand the Gemma family. Earlier, the company released more powerful versions, such as Gemma 4 12B, which support multimodal input and audio, as well as quantized versions to reduce memory requirements. This makes the technology accessible to a broader range of devices.