Source: aws.amazon.com/blogs
In early December, Amazon Web Services hosted its premier conference in Las Vegas — AWS re:Invent 2025. This year’s event featured an unprecedented number of technological premieres. In this article, check out the event’s key announcements that will help you build, scale, and transform your business in the cloud.
Top AWS re:Invent 2025 Announcements
Analytics
AWS Clean Rooms launches privacy-enhancing synthetic dataset generation for ML model training
Train ML models on sensitive collaborative data by generating synthetic datasets that retain statistical patterns while protecting individual privacy with configurable noise levels and re-identification protection.
Artificial Intelligence
Amazon Nova 2 Sonic: A new speech-to-speech model for conversational AI
Create more natural voice interactions with improved speech-to-speech AI — now supporting multilingual conversations, dynamic speech steering, cross-modal inputs, and improved telephony integration while maintaining conversational context across tasks.
Amazon Nova 2 Lite: A fast and cost-effective model for generating quality results
Nova 2 Lite offers fast, cost-effective AI for everyday tasks. It features advanced reasoning capabilities, built-in tools, and a one-million-token context window.
Amazon Nova Forge: Build your own frontier models using Nova
The new program gives organizations unprecedented access to Nova model training so they can build their own frontier models with deep industry expertise integration, without the traditional barriers of cost, compute, and time.
Amazon Nova 2 Omni in preview
A universal model for multimodal reasoning and image generation that supports text, image, video, and audio input, generating both text and image outputs.
Build reliable AI agents for UI workflow automation with Amazon Nova Act, now generally available
A new AWS service helps developers build AI agents that automate in-browser tasks like form filling, data search and extraction, shopping, and booking, as well as QA testing — with over 90% reliability for enterprise adoption.
Amazon Bedrock AgentCore adds quality evaluations and policy controls for deploying trusted AI agents
New capabilities help companies deploy agents with strengthened policy controls, quality monitoring, improved memory, and natural conversation ability, making it easier to confidently scale AI solutions across the organization.
Amazon S3 Vectors is now generally available with increased scale and performance
Take vector storage and query scaling to new heights with the general availability of S3 Vectors — now supporting up to 2 billion vectors per index, 100ms query latency, and expanded regional availability, all while reducing costs by up to 90% compared to specialized databases.
Access fully managed foundation models from leading providers like Google, Kimi AI, MiniMax AI, Mistral AI, NVIDIA, OpenAI, and Qwen, including the new Mistral Large 3 and Ministral 3 (3B, 8B, and 14B) models via Amazon Bedrock.
Accelerate AI development using Amazon SageMaker AI with serverless MLflow
Simplify AI experimentation with infrastructure-free MLflow that launches in minutes, scales automatically, and integrates seamlessly with SageMaker model tuning and pipeline capabilities.
Amazon Bedrock adds reinforcement fine-tuning, making it easier for developers to build smarter and more accurate AI models
A new model tuning capability uses feedback-based learning, delivering a 66% accuracy boost over baseline models without the need for large labeled datasets or deep ML expertise.
Checkpointless and elastic training on Amazon SageMaker HyperPod
Accelerate AI model development with new training features that provide instant recovery from failures and auto-scaling based on resource availability.
New serverless customization in Amazon SageMaker AI accelerates model fine-tuning
New customization provides an easy-to-use interface for the latest fine-tuning techniques, speeding up the AI model tuning process from months to days.
Compute
AWS introduces Graviton5 — the most powerful and efficient AWS processor
The fifth-generation chip delivers the best price-performance for a wide range of workloads on Amazon EC2.
Trainium3 UltraServers now available: Enabling customers to train and deploy AI models faster and at lower cost
Amazon EC2 Trn3 UltraServers, powered by the first 3nm AI chip from AWS, help organizations of all sizes tackle their most ambitious AI training and inference tasks.
Amazon EC2 X8aedz instances powered by 5th Gen AMD EPYC processors for memory-intensive workloads
New memory-optimized instances deliver CPU frequencies up to 5 GHz and 3 TiB of memory — ideal for electronic design automation and memory-intensive databases requiring high single-thread performance.
AWS Lambda Managed Instances: Serverless simplicity with EC2 flexibility
Run Lambda functions on EC2 compute while retaining serverless simplicity — unlocking access to specialized hardware and cost optimization via EC2 pricing models, while AWS handles all infrastructure management.
Build multi-step applications and AI workflows with AWS Lambda Durable Functions
A new Lambda capability allows you to build applications that reliably coordinate multiple steps over long periods — from seconds to a year — without paying for compute idle time while waiting for external events or human decisions.
Wise IT is an official AWS Advanced Partner in Ukraine! Book a free consultation on cloud migration and service implementation:
Containers
Announcing Amazon EKS capabilities for workload orchestration and cloud resource management
Streamline Kubernetes development with fully managed platform capabilities that handle workload orchestration and cloud resource management, eliminating infrastructure maintenance needs and ensuring enterprise-grade reliability and security.
Databases
Database Savings Plans for AWS databases
A new pricing model helps maintain cost efficiency by providing flexibility in database service selection and deployment options. New capabilities optimize costs and improve scalability for Amazon RDS for SQL Server and Oracle
Manage development, test, and production databases more efficiently with new features, including Developer Edition support for SQL Server, CPU-optimized M7i/R7i instance support, and expanded storage options up to 256 TiB.
Amazon OpenSearch Service improves vector database performance and cost with GPU acceleration and auto-optimization
Build and optimize large-scale vector databases up to 10x faster and at one-quarter the cost with new GPU acceleration and auto-optimization capabilities that automatically balance search quality, speed, and resource usage.
Global Infrastructure
AWS AI Factories
Deploy fully managed AWS AI infrastructure in your data centers with pre-integrated foundation models, specialized hardware, and AWS services, helping enterprises and government agencies accelerate AI initiatives while meeting data residency requirements.
Management and Governance
AWS DevOps Agent helps accelerate incident response and improve system reliability (preview)
Acting as an autonomous on-call engineer, AWS DevOps Agent analyzes data from CloudWatch, GitHub, ServiceNow, and other tools to identify root causes and coordinate incident response.
New and enhanced AWS Support plans add AI capabilities to expert guidance
Prevent cloud infrastructure issues before they impact your business with AWS Support plans that combine AI-powered insights with expert advice, offering faster response times and proactive monitoring of performance, security, and costs.
Amazon CloudWatch introduces unified data management and analytics for operations, security, and compliance
Reduce data management complexity and costs with automated normalization from disparate sources, native analytics integration, and built-in support for industry-standard formats like OCSF and Apache Iceberg.
Migration and Modernization
AWS Transform custom: Crush technical debt with AI-powered code modernization
Modernize enterprise applications at scale with this new AI-powered service that learns your organization’s patterns, automates transformations across repositories, and reduces turnaround time by up to 80% with both ready-made and custom capabilities.AWS Transform announces full-stack Windows modernization capabilities
Modernize complex Windows applications up to five times faster with coordinated AI-powered transformation covering application code, UI frameworks, databases, and deployment configurations.
AWS Transform for Mainframe introduces Reimagine capabilities and automated testing functionality
New AI-powered capabilities help transform legacy mainframe applications into cloud architectures by automating complex testing processes — reducing modernization timelines from years to months through intelligent analysis and automated test generation.
implementation case study by Wise IT
Networking and Content Delivery
Amazon Route 53 Global Resolver for secure anycast DNS resolution (preview)
Simplify hybrid DNS management with a single service that resolves public and private domains globally via a secure anycast connection, reducing operational costs and maintaining consistent security controls.
Partner Network
AWS Partner Central is now available in the AWS Management Console
Access Partner Central directly via the AWS Console to streamline your journey from customer to Partner — manage solutions, opportunities, and marketplace listings in a single unified interface with enterprise-grade security.
Security, Identity, and Compliance
New AWS Security Agent proactively secures applications from design to deployment (preview)
Scale your AppSec expertise with AI-powered design reviews, code analysis, and contextual penetration testing that factor in your unique security requirements and application architecture.
Amazon GuardDuty adds extended threat detection for Amazon EC2 and Amazon ECS
Extended threat detection now provides unified visibility across VM and container environments, helping teams identify complex multi-stage attacks spanning different AWS workloads supporting the same application.
AWS Security Hub is now generally available with near real-time analytics and risk prioritization
Security Hub helps teams respond to risks faster by correlating security signals in near real-time across AWS environments. Simplify IAM policy creation with IAM Policy Autopilot, a new open-source MCP server for builders
Accelerate AWS development with an open-source tool that analyzes your code to generate valid IAM policies, providing AI assistants with up-to-date AWS service knowledge and reliable permission recommendations.
Storage
Amazon FSx for NetApp ONTAP now integrates with Amazon S3 for seamless data access
Amazon FSx for NetApp ONTAP now allows file system data access via Amazon S3, enabling seamless integration with AWS analytics, ML, and generative AI services without moving or copying data.
Announcing replication support and Intelligent-Tiering for Amazon S3 Tables
New features ensure automatic cost optimization via intelligent storage tiering and simplified table replication across AWS regions and accounts.
Amazon S3 Storage Lens adds performance metrics support for billions of prefixes and export to S3 Tables
New capabilities help optimize application performance, analyze unlimited prefixes, and simplify metrics analysis via integration with S3 Tables.
Wise IT is an official AWS Advanced Partner in Ukraine! Book a free consultation on cloud migration and service implementation: