Google AI Updates: April 7, 2026

1. Google Launches Veo 3.1 Lite, Cuts Video Generation Costs by 50%+

Google. Google released Veo 3.1 Lite, a cost-optimized video generation model priced at $0.05/second for 720p and $0.08/second for 1080p — less than half the cost of Veo 3.1 Fast with no speed penalty. The model supports text-to-video and image-to-video, landscape and portrait ratios, and 4/6/8 second durations. Available through the Gemini API and Google AI Studio. Source

2. Google Releases AI Edge Eloquent, an Offline-First Dictation App on iOS

Google. Google quietly released “AI Edge Eloquent,” a free offline-first AI dictation app for iOS that uses on-device Gemma-based ASR models to transform raw speech into polished text. The app features automatic filler-word removal, text transformation options (Key Points, Formal, Short/Long), a fully offline mode where all audio stays on-device, and optional Gemini integration for enhanced refinement. It also includes a Dictionaries feature that can import jargon from Gmail. Source

3. Certified Kubernetes AI Conformance Program Standardizes GPU/TPU Workload Management

Google. Google published a deep-dive on the Certified Kubernetes AI Conformance program, a cross-industry standard co-led by Google’s Janet Kuo alongside engineers from Microsoft, Red Hat, and Kubermatic. The program adds four AI-specific pillars to standard K8s conformance: Dynamic Resource Allocation for fine-grained GPU/TPU control, all-or-nothing scheduling to prevent distributed training deadlocks, intelligent autoscaling based on custom AI metrics, and standardized observability for accelerator performance monitoring. Source