
Google’s upcoming Pixel 10 and Pixel 11 are generating significant buzz, with recent leaks hinting at a slew of impressive camera and AI capabilities. While hardware specifications often surface months in advance, software features tend to remain under wraps until closer to launch. However, a recent leak from Google’s gChips division has provided insights into the exciting new AI tricks and software-based camera features these devices may offer.
AI-Powered Enhancements
The Tensor G5 chip, expected to debut in the Pixel 10 series, will reportedly feature “Video Generative ML” capabilities, leveraging AI for intuitive video editing within the Photos app and potentially YouTube Shorts. Further enhancing the user experience, Pixel 10 and 11 may incorporate “Speak-to-Tweak,” an LLM-based editing tool, and “Sketch-to-Image,” allowing users to generate images from sketches, similar to a feature offered by Samsung’s Galaxy AI. Additionally, the Tensor G5 is rumored to support on-device Stable Diffusion models for AI image generation within the Pixel Studio app.
Camera Advancements
Google has consistently prioritized camera quality in its Pixel phones, and the Pixel 10 and 11 appear to be no exception. The Tensor G5 will finally support 4K 60fps HDR video recording, a significant upgrade from the previous 4K 30fps limitation. Looking ahead to the Pixel 11, leaks suggest the possibility of 100x zoom capabilities for both photos and videos, powered by Machine Learning and a “next-gen” telephoto lens.
The Pixel 11 may also introduce “Ultra Low Light video,” an on-device version of the existing Night Sight feature, enabling high-quality video capture in extremely low-light conditions (5-10 lux). Furthermore, Cinematic Blur on the Pixel 11 is expected to receive an upgrade with 4K 30fps support and a new “video relight” feature for adjusting lighting conditions.
Ambient Computing and Health Tracking
With the introduction of a new “nanoTPU” in the Tensor G6, slated for the Pixel 11, Google is reportedly focusing on ML-based always-on features. These include health-related functions such as agonal breathing detection, cough and snore detection, sleep apnea monitoring, fall detection, gait analysis, sleep stage monitoring, and emergency sound detection. Activity tracking features like “Running ML” may also be included, offering tools like “coachable pace” and “balance & oscillation” analysis.