HeyGen, an AI video generation platform, announced that its Avatar IV diffusion model, which runs on more than 18B parameters, is now 1.86× faster on Google Cloud's eight-chip Trillium (v6e) TPU host …
Google has released Credentio, an open-source C++ library for validating C2PA Content Credentials, supporting specification versions 2.2 and 2.4, and already powering nearly 40 conformant Google produ…
Google's Go programming language, created by Rob Pike, Robert Griesemer, and Ken Thompson, is positioned as ideal for AI-assisted software engineering because it prioritizes code review, verification,…
Google AI Edge's LiteRT and LiteRT-LM enable deployment of Gemma 4 E2B on Raspberry Pi 5, achieving 99 tokens/sec prefill and 9 tokens/sec decode with a peak memory footprint of 1432 MB, powering the …
Google is joining the Agent Plugins project as a Core Maintainer, represented by Kevin Hou, and will integrate the open, vendor-neutral specification into its products. The Agent Plugins 1.0.0 specifi…
Google, Hugging Face, and other industry partners released the 2026-07-28 Model Context Protocol specification candidate, which removes transport-level session management to make the protocol stateles…
Google Cloud API Gateway has introduced model routing in Public Preview, enabling developers to route OpenAI-compatible requests to Gemini, Claude, or OpenAI OSS-GPT models via a serverless ingress la…
LiveKit, a real-time AI infrastructure provider, reports that traditional load balancing metrics like QPS and CPU utilization fail for real-time AI agents, which require session-aware balancing that t…
Google Cloud announced that Agent and Model Evaluations in its Gemini Enterprise Agent Platform are now generally available, enabling developers to measure and compare agents and models during develop…
Google released a microbenchmark suite on GitHub to help developers evaluate TPU performance by measuring network, compute, HBM, and host transfer capabilities. The suite establishes a Speed-of-Light …
Ray AI libraries (Serve, Data, Train) now support Google TPU slices through a topology field that reserves a whole ICI-connected slice, preventing multi-host deployment hangs. Ray Serve serves LLMs vi…
Google's Tunix post-training library introduces a high-throughput framework for agentic reinforcement learning that keeps TPUs fully utilized during multi-turn training. Tunix uses an asynchronous tra…
Google announced that Android apps can integrate with the platform's intelligence system using AppFunctions, enabling voice and text commands to perform tasks like expense tracking, itinerary manageme…
Google's Firebase AI Logic SDK now supports cloud and hybrid inference for Android apps, enabling developers to build intelligent features that combine on-device Gemini Nano models with cloud fallback…
Google has open-sourced Jetpacker, a demo travel app built for Google I/O that showcases how to integrate generative AI features into Android apps using on-device and cloud models. The app uses Gemini…
Google announced that its Gemini Nano 4 model, now running on over 140 million devices, can be used through ML Kit's Prompt API to build on-device intelligent features in Android apps, such as summari…
Ray 2.55 makes Google Cloud TPUs a first-class accelerator, with official pre-built images and support across core libraries. The distributed-computing framework now treats TPU slices as schedulable r…
Google has evolved its Conductor tool from a Gemini CLI extension into a portable plugin, now supporting Antigravity CLI and other tools, enabling conversational spec-driven development with persisten…
Modular prompt transpilation, treating prompts as build artifacts with deterministic builds and CI/CD integration, solves the scaling problems of monolithic system prompts for AI agents, according to …
Google Cloud announced that Parallel Web Systems is now a natively integrated web grounding provider on the Gemini Enterprise Agent Platform, enabling developers to ground Gemini models in real-time w…