Google introduces TorchTPU, a native integration between PyTorch and TPUs, enabling developers to easily migrate their models to Google’s custom chips with a focus on ease of use, portability, and maximum performance in distributed hyperscale AI systems.
MaxText now supports Supervised Fine-Tuning and Reinforcement Learning on single-host TPUs, enabling developers to optimize their language models with modern post-training techniques.
Gemini CLI now supports specialized subagents that handle complex tasks in parallel, each with its own tools and context window, making the main session faster and more efficient.
Agents CLI unifies AI agent development, evaluation and deployment on Google Cloud, dramatically reducing time to production and enabling AI coding assistants to seamlessly leverage the entire infrastructure.
Google’s ADK enables AI agents to dynamically load specialized domain expertise and independently generate new expertise areas without overwhelming system prompts.
ADK Go 1.0 enables secure, traceable multi-agent systems in Go through OpenTelemetry integration, plugin architecture, and human-in-the-loop approvals.
Google introduces TorchTPU, a solution enabling native PyTorch on TPUs, allowing developers to migrate their ML workloads to Google’s supercomputing infrastructure with minimal code changes and full utilization of TPU resources.
Google demonstrates how to build long-lived AI agents with the Agent Development Kit that work for weeks, pause, and retain their memory – without the context loss of traditional chatbots.
Google’s new Gemini Embedding 2 unifies text, images, video, audio and documents in a single embedding model, supporting over 100 languages and processing multiple media types simultaneously to enable new multimodal AI applications and intelligent search systems.
Google Genkit introduces a new middleware system for extending and securing AI applications, with modular hooks enabling retries, fallbacks, and human oversight—available in TypeScript, Go, and Dart, with Python support coming soon.
Google extends MaxText with Supervised Fine-Tuning and Reinforcement Learning for single-host TPUs, enabling efficient post-training of language models on v5p-8 and v6e-143 systems.
Google enhances the Google Pay API with new features for merchant-initiated payments, enabling better control over subscriptions, deferred payments and auto-reloads, plus increased transparency for users.
