11/20/2026, 12:00:00 AM ~ 11/23/2026, 12:00:00 AM (UTC)
Recent Announcements
Today, Amazon CloudWatch announces support for your own application telemetry in the CloudWatch Observability EKS add-on. The CloudWatch agent you already run now accepts metrics, traces, and logs from your applications, so you don’t deploy and operate a separate collector to get them into CloudWatch.\n Applications instrumented with an OpenTelemetry SDK send to an OTLP endpoint that the agent serves on every node, on by default. Metrics arrive as CloudWatch metrics you query with PromQL, traces go to AWS X-Ray and Transaction Search, and logs go to CloudWatch Logs.
For Prometheus metrics, the agent scrapes the targets described by standard ServiceMonitor and PodMonitor objects, with no Prometheus operator to run and no change to your scrape configuration. Each team defines its own targets alongside its application rather than in one central add-on configuration. Label selectors let you choose which monitors CloudWatch scrapes.
Both sources carry the same Kubernetes, host, cloud, and service attributes the add-on applies to the pod they came from, so you can join application telemetry to your cluster metrics.
Custom telemetry is available in version 8.0.0 and later of the EKS add-on and Helm chart, in all AWS Regions where CloudWatch supports OpenTelemetry metrics. Telemetry is charged at the Amazon EKS OpenTelemetry ingestion rate of $0.08 per GB. For details, see Amazon CloudWatch insights pricing.
To get started, see Custom telemetry in the Amazon CloudWatch User Guide.
Amazon Transcribe announces a synchronous transcription API
Amazon Transcribe now offers a synchronous transcription API,
TranscribeAudio. Send a complete audio file in a single request and get the full transcript back in the same response. There is no job to start, no status to poll, no connection to manage. Add fast, accurate speech-to-text to interactive applications such as voice assistants, agent-assist tools and on-demand “transcribe this clip” features.\n The synchronous API supports the Amazon Transcribe features you already use, including:
Word-level timestamps and smart formatting for numbers, dates, currency and phone numbers
Automatic language identification, including audio with more than one language
Speaker diarization, and channel identification for up to two channels
Custom vocabulary and vocabulary filtering, which you can provide directly in the request with no setup step
PII redaction
Send audio inline (up to 25 MB) or from Amazon S3 (up to 500 MB), 0.5 seconds to 1 hour long, in WAV, MP3, FLAC, M4A, MP4, Ogg, WebM or AMR format. The transcript comes back in the response, or Transcribe can write it to an S3 location you choose. Pricing is $0.0070 per minute of audio, billed per second, with speaker diarization, language identification, custom vocabulary and channel identification included. Two-channel audio is billed once on file duration. PII redaction is available as an add-on.
Supports English, Spanish, French, Italian, German, Japanese, Portuguese, Korean and Chinese (Simplified, Traditional and Hong Kong). Available in US East (N. Virginia, Ohio), US West (Oregon), Canada (Central), Europe (Ireland, London, Frankfurt), Asia Pacific (Mumbai, Singapore, Sydney, Tokyo, Seoul) and South America (São Paulo), with PII redaction available in all stated Regions except South America (São Paulo).
To get started, see the Amazon Transcribe Sync API launch blog, Amazon Transcribe API Reference and the Amazon Transcribe pricing page.
Amazon DynamoDB vector search now supports new inline filter operators
Amazon DynamoDB vector search inline filters now support range comparisons, value lists, and logical operators in addition to the equality operator, so you can narrow a vector search to exactly the items you want. You can now filter on numeric ranges such as price, rating, or publish date, match an attribute against a list of values, check whether an attribute exists, and combine conditions with OR and NOT. Amazon DynamoDB vector indexes store vector embeddings alongside your operational data, and the SearchVectors API returns the items whose embeddings are most similar to your query vector.\n The search condition expression in a SearchVectors request now accepts the comparison operators <, <=, >, >=, and BETWEEN on Number attributes, the IN operator, the attribute_exists and attribute_not_exists functions, and the logical operators AND, OR, and NOT. DynamoDB applies the filter during the search, so every result it returns matches. For example, a retail application can return the products most similar to a shopper’s query within a price range and a set of brands, and a retrieval-augmented generation application can limit results to selected document categories published after a given date.
The new operators are available in all AWS Regions where DynamoDB vector search is available.
To learn more, see vector search in the Amazon DynamoDB Developer Guide, or visit the Amazon DynamoDB page.
Amazon SageMaker JumpStart now offers a model leaderboard in Amazon SageMaker Studio. You can use it to compare foundation models side by side on throughput, latency, and cost, and then deploy or customize the model that fits your workload.\n Comparing models has usually meant reading model cards one at a time and relying on benchmark scores that different sources published under different conditions. The JumpStart leaderboard puts consistent measurements in one view. AWS evaluates every model on three reasoning benchmarks, BIG-Bench Hard, GPQA Diamond, and MMLU-Pro, using the same evaluation setup for each model. AWS also measures time to first token, throughput, and instance cost on SageMaker for each JumpStart deployment configuration. You can use all of the filters already available for JumpStart models, such as model provider, license, and input and output modality, to narrow the list. You can also choose a use case profile and an optimization goal (balanced, cost, latency, or throughput) to see the metrics for the configuration you plan to deploy. At launch, the leaderboard covers more than 70 LLMs from providers including Meta, Google, Qwen, IBM, and Arcee.
The JumpStart model leaderboard is available at no additional charge in all AWS commercial Regions where Amazon SageMaker Studio is available. To get started, open SageMaker JumpStart in SageMaker Studio and choose the Leaderboard view. To learn more, see JumpStart leaderboard methodology and read the launch blog post.
AWS Partner Central Agents now help AWS Partners create and manage AWS Partner Central and AWS Marketplace support cases in one place. The agents are available 24/7 and draw on Partner Central account details and authoritative, partner-specific knowledge from AWS to resolve most inquiries instantly. If a inquiry needs a support engineer, the agents will gather details and create a case. This reduces the time partners spend looking for help, and gives support teams the context they need to resolve cases faster.\n Partners can handle a case from start to finish in one chat, with the full record in AWS Support Center. To create a case, partners can start describing their issue to AWS Partner Central Agents or use the prompt “get support.” The agents gather additional details necessary to resolve the case, set the category, and draft the description. Partners review and submit the case, and the agents route the case to the right support team. In the same chat, partners can view case communications, track progress, post replies, and add attachments. With this launch, Partner Central and AWS Marketplace cases also move to AWS Support Center. There, partners can view the full case history and translate support correspondence into Chinese, French, German, Japanese, Korean, Portuguese, and Spanish.
The new support experience is available today to all AWS Partners and AWS Marketplace sellers in the AWS Partner Central console and through the AWS Partner Central Agents MCP server. With the MCP server, partners can bring the support experience into tools such as Kiro, AWS Quick, CRM systems, and other AI clients. To learn more, see Getting support in AWS Partner Central, Getting support for AWS Marketplace, and Getting started with the Partner Central Agents MCP server.
Spend limits in the new AWS experience now support SMS alerts
Builders using the new AWS experience on the Paid plan can now receive spend limit alerts by text message (SMS), alongside existing email notifications. A spend limit sets the maximum monthly cost for a project. Notifications are sent as a project approaches its limit, and when a project reaches its limit, AWS automatically pauses it. SMS gives builders a second way to stay informed about their project costs.\n Email alerts can go unnoticed when builders are away from their inbox, leaving little time to respond before a project is paused. SMS alerts close this gap with a text message when a project’s actual costs reach 90% of its spend limit, so builders have time to act, such as raising the limit or stopping resources they don’t need.
To turn on SMS alerts, verify your mobile phone number in AWS Settings and turn on text alerts for spend limits. Text alerts are only sent to verified numbers, and you can opt out at any time. SMS alerts for spend limits are available at no additional charge for US phone numbers, including Puerto Rico, Guam, the US Virgin Islands, and American Samoa.
To learn more, see Create a spend limit in AWS Settings, the AWS Account Management Reference Guide, or Get started with AWS.
Amazon S3 Vectors raises the filterable metadata limit to 40 KB per vector
Amazon S3 Vectors now supports up to 40 KB of filterable metadata per vector, up from 2 KB. You can attach and filter on more attributes per vector, such as category, permissions, timestamps, and descriptive tags, so your retrieval-augmented generation (RAG), agentic, and semantic-search applications can narrow results by more of the fields that matter to them.\n S3 Vectors provides native support to store and query vectors in Amazon S3, delivering purpose-built, cost-optimized vector storage and query at billion-vector scale. With this launch, indexes in new vector buckets support the higher limit by default, with no change to how you write vectors with PutVectors or run filtered queries with QueryVectors. To use the higher limit on an existing index, update it in place to enhanced index mode with the UpdateIndexMode API.
The higher filterable metadata limit is available at no additional cost in all AWS Regions where Amazon S3 Vectors is available. To get started, use the AWS CLI, AWS SDKs, or the Amazon S3 console. To learn more, visit the Amazon S3 Vectors documentation.
AWS Glue Data Catalog introduces data quality monitoring for catalogs and databases
AWS Glue Data Catalog introduces data quality monitoring for catalogs and databases, powered by AWS Glue Data Quality, which profiles and evaluates your tables at catalog scale from a single configuration. You define data quality checks once for a catalog, a database, or a single table, and AWS Glue Data Quality evaluates every table in scope, including tables added later, on the cadence you choose. Quality issues surface as data changes, without per-table setup or manual runs, which helps keep bad data from reaching downstream consumers.\n You can use built-in profiling options that AWS Glue provides for common baseline statistics or provide your own Data Quality Definition Language (DQDL) ruleset with analyzers, rules, or both to hold every table in a database to the same quality standards. You have the flexibility to run evaluations once when you create the configuration or set up triggers, so evaluations run on a schedule or whenever table metadata changes, including every Apache Iceberg commit. After three runs on a table, AWS Glue Data Quality flags anomalous statistics that drift outside expected bounds, surfacing changes that no rule covers. Results are accessible through the existing AWS Glue Data Quality APIs, or you can opt-in to write them to Apache Iceberg tables in the AWS Glue Data Catalog or to Amazon S3.
Data quality monitoring for catalogs and databases in AWS Glue Data Catalog is available in all AWS Commercial, AWS GovCloud (US), and AWS China regions.
To get started, create a data quality configuration by using the AWS CLI, API, or SDKs, which support built-in profiling and custom DQDL rulesets at the catalog, database, or table level. In Amazon SageMaker Unified Studio, you can turn on profiling for a catalog or database from the catalog page. To learn more, visit the AWS Glue Data Quality documentation and the Amazon SageMaker Unified Studio documentation.
AWS Pricing Calculator now estimates workload costs through a new API
AWS Pricing Calculator now offers GetUsageLinesEstimatedCost, a new API that returns the estimated cost of a planned workload in a single API call. You send up to 500 usage lines in one request and get the cost back right away, without saving an estimate to your account. While workload estimates let you create and save an estimate in Pricing Calculator, GetUsageLinesEstimatedCost returns a cost on demand, directly in your own tools and workflows.\n You can choose from three rate types: public On-Demand rates, your rates after discounts, or your rates after discounts plus Savings Plans and Reserved Instances. The API returns the cost of each line and the total. For example, you can check what a design will cost before you deploy it, or let an AI coding agent call the API to price a workload while it helps you design it.
GetUsageLinesEstimatedCost is generally available at no additional charge in all AWS Regions, except the AWS GovCloud (US) Regions and the China Regions.
To learn more, see GetUsageLinesEstimatedCost in the AWS Billing and Cost Management API Reference, or visit the AWS Pricing Calculator product page.
AWS Billing and Cost Management now offers APIs to track payment status
Today, AWS Billing and Cost Management announces new APIs that let you view your AWS payment information through the AWS SDKs and the AWS CLI. With these APIs, you can check your account balance, find invoices that are due or past due, see which payments and credit memos were applied to each invoice, and confirm that the payment methods on your account are active. You can call the APIs from your own tools or through an AI agent.\n Previously, finance teams had to sign in to the console to find past-due invoices, match payments to invoices, or check whether a payment method was still active. Now you can automate these checks with APIs such as GetFinancialRecord, BatchGetFinancialRecord, ListFinancialRecords, GetAccountBalance, GetPaymentInstrument and ListPaymentInstruments. You can integrate the APIs with your accounts payable systems to monitor past-due balances and reconcile AWS invoices on your own schedule.
The APIs are available in the US East (N. Virginia) Region, and available to customers in all commercial AWS Regions.
To learn more, visit the AWS Payments API Reference.
Amazon EventBridge Scheduler adds 589 new API actions and event bus targets
Amazon EventBridge Scheduler expands its AWS SDK integrations with 589 new API actions across 81 AWS services, and adds templated targets for the enhanced EventBridge Custom event bus. You can now schedule direct invocations of a broader set of AWS services without writing custom integration code.\n EventBridge Scheduler is a serverless scheduler that allows you to create, run, and manage billions of scheduled events and tasks across more than 270 AWS services, without provisioning or managing the underlying infrastructure. With a new templated target, you can schedule event delivery directly to the Custom event bus using its PutEvents or PutRawEvents API. The templated target also makes it easier to add metadata fields to each event. For example, you can send an event to your bus every night at midnight to start a reconciliation workflow, with no function in between to call the API.
The new API actions are available in all commercial AWS Regions where Amazon EventBridge Scheduler is available, subject to the availability of each target service in the Region. The templated targets for the Custom event bus are available in all AWS Regions where the enhanced Custom event bus is available. To learn more, see Using templated targets in EventBridge Scheduler and Using universal targets in EventBridge Scheduler in the Amazon EventBridge Scheduler User Guide. Install the Agent Toolkit for AWS to start building with your AI coding agent.
AWS CloudFormation Service-Managed StackSets adds custom execution roles with nested stacks support
AWS CloudFormation StackSets now supports custom execution roles for service-managed stack sets, which enables you to scope the permissions StackSets uses in each target AWS account. StackSets deploys the same infrastructure, such as networking and security settings, across thousands of accounts and Regions in one operation. With a custom execution role, you can also use nested stacks, macros, and transforms in service-managed stack sets, which allows you to deploy the reusable templates you use for individual stacks across your organization.\n With a custom execution role, you define what StackSets can do in each account. By default, service-managed stack sets use a role in each account that can provision any resource in a template without additional setup, and that default remains available. Now, you can provide a template that defines an IAM role, called the execution role, and StackSets creates it in each account before deploying. For example, a stack set that creates Amazon S3 buckets can use an execution role limited to CloudFormation and Amazon S3 permissions. When you update the execution role template, StackSets updates the role in every account it deployed to.
To get started, choose Set execution role when you create a stack set in the CloudFormation console, or use the AWS CLI or AWS SDKs. To learn more, visit the documentation and the AWS DevOps Blog. This feature is available at no additional charge in all commercial AWS Regions, the AWS GovCloud (US) Regions, and the China Regions. Refer to the AWS Region table for details.
Amazon EventBridge Custom event bus now supports Webhook Events
Amazon EventBridge now supports Webhook Events for the enhanced Custom event bus, so SaaS applications, AI agents, and custom integrations can send events to a bus over HTTPS. You no longer need to deploy and operate your own API endpoint and compute to receive webhooks and publish them to your bus.\n The EventBridge Custom event bus is a serverless event broker that enables you to create scalable event-driven applications by routing events between your applications, SaaS integrations, and AWS services. When you create a webhook event source on your bus, EventBridge provides a managed HTTPS URL and verifies each request using an HMAC signature, a static token, or basic authentication. EventBridge then publishes accepted events to the bus, where Subscribers filter and deliver them like any other event. For example, shipment status updates from a logistics partner can reach your order tracking and customer notification Subscribers as soon as the partner sends them.
Webhook Events are available in all AWS Regions where the enhanced Custom event bus is available. Events received through a webhook are billed as standard Custom event bus ingress, with no additional charge for the webhook.
To learn more, see Webhook Events in the Amazon EventBridge User Guide and the Amazon EventBridge product page. Install the Agent Toolkit for AWS to start building with your AI coding agent.
Amazon EventBridge Custom event bus now supports Amazon MSK as a target
Amazon EventBridge now supports Amazon Managed Streaming for Apache Kafka (Amazon MSK) as a target for the enhanced Custom event bus, so you can deliver events to Kafka topics without running your own producer.\n The EventBridge Custom event bus is a serverless event broker that enables you to create scalable event-driven applications by routing events between your applications, SaaS integrations, and AWS services. To deliver events to MSK, you create a Subscriber that specifies the cluster ARN and topic. You can also configure Kafka producer settings such as partition key, acknowledgment level, and record headers. EventBridge supports both public and private, serverless and provisioned clusters. When you create a Subscriber for a private cluster, EventBridge automatically sets up a private connection to the cluster. For example, teams that publish to a shared Custom event bus can now deliver purchase-order events to the MSK topics their payments team already consumes from.
Amazon MSK targets are available in all AWS Regions where the enhanced Custom event bus is available. Delivery to MSK is billed as standard Custom event bus egress. EventBridge also supports compression before delivery, which is billed as an Event Evaluation.
To learn more, see Amazon MSK cluster target in the Amazon EventBridge User Guide and the Amazon EventBridge product page. Install the Agent Toolkit for AWS to start building with your AI coding agent.
Amazon Aurora MySQL 8.4.12 is now generally available
Starting today, Amazon Aurora MySQL-Compatible Edition 8.4 supports MySQL 8.4.12. This release includes several security enhancements and bug fixes. We recommend upgrading to the newer minor versions to accept fixes for Common Vulnerabilities and Exposures (CVEs) in prior versions of MySQL and to benefit from bug fixes, performance improvements, and new functionality added by community MySQL. For more details, refer to the Aurora MySQL 8.4 and MySQL 8.4.12 release notes.\n You can upgrade your databases during scheduled maintenance windows using automatic minor version upgrades. To simplify operations at scale, enable automatic minor version upgrades and use the AWS Organizations Upgrade Rollout Policy to orchestrate upgrades across your clusters in phases, validating on lower-priority environments before upgrading your most critical ones. You can perform minor version upgrades in-place or via snapshot restore. Learn more about performing version upgrades in the Amazon Aurora User Guide. This release is supported in all AWS Regions where Aurora MySQL is available.
Amazon Aurora is designed for high performance and availability at global scale with full MySQL compatibility. It provides scale-to-zero serverless compute, Aurora Global Database for multi-Region resilience, Aurora I/O-Optimized for improved price performance on I/O-intensive workloads, and built-in security and continuous backups. To get started, take a look at Aurora’s getting started page.
Amazon Connect Customer expands how teams track and act on service level agreements for cases
Amazon Connect Customer enhances service level agreements (SLAs) for cases, so teams can set service goals for more of their work and respond faster when a case falls behind. Admins can set SLAs on case fields beyond status, such as requiring a case’s severity to drop from high to medium within 24 hours. Admins can add an email action to SLA breach rules, so the assigned agent learns right away when a case misses its SLA. Developers can use APIs to adjust a case’s SLA deadline when the team is waiting on information from an outside party, such as a partner.\n Cases is available in the following AWS Regions: US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), and Africa (Cape Town). To learn more and get started, visit the Cases webpage and documentation.
AWS Transform announces UI component modernization for Windows web applications
Today, AWS announced the general availability of third-party UI component porting for ASP.NET web site modernization in AWS Transform. When AWS Transform modernizes a legacy ASP.NET web site, it now detects dependencies on popular third-party UI component libraries. You can replace them with license-free Microsoft-native and open-source components, or you can keep using vendor UI libraries.\n AWS Transform rewrites legacy ASP.NET web sites for ASP.NET Core UI frameworks. When the AWS Transform for .NET modernization agent finds a dependency on a popular third-party UI library, it offers two options. The first option is for customers who require open-source software only. The agent replaces each proprietary control with a Microsoft-native or open-source control on the target UI framework and removes the original dependency. The second option is for customers who want to keep using vendor libraries. The agent adds code comments and next-step instructions for moving to an equivalent vendor library on the target UI framework. You then arrange licensing with the vendor and finish the port using an AI coding companion.
This capability is available at no additional cost in all eight AWS Regions where AWS Transform is offered: US East (N. Virginia), Canada (Central), Asia Pacific (Mumbai), Asia Pacific (Seoul), Asia Pacific (Sydney), Asia Pacific (Tokyo), Europe (Frankfurt), and Europe (London).
To learn more, see Porting UI layer code in the AWS Transform User Guide.
AWS Transform announces support for .NET 11 and expanded project types
Today, AWS announced the general availability of .NET 11 support in AWS Transform for .NET. Support for VB.NET, desktop projects, and mobile projects, previously in preview, is also now generally available. You can modernize a broad range of Windows .NET workloads to the latest version of .NET and lower costs by running them on Linux. Modernized workloads gain the performance improvements, new C# language features, and improved diagnostics in .NET 11.\n AWS Transform modernizes both legacy .NET Framework applications (version 3.5 and later) and modern .NET applications (versions 5 through 10) to .NET 11. It transforms code written in C# and VB.NET. Supported project types include class libraries, console applications, ASP.NET web projects (MVC, Web API, and Web Forms), web services (ASMX and WCF), desktop applications (WinForms and WPF), mobile applications (Xamarin), unit test projects (MSTest, NUnit, and xUnit), and Windows services. You can run all transformed applications except desktop and mobile projects on Linux, which can reduce operating costs by up to 40%.
This capability is available at no additional cost in all eight AWS Regions where AWS Transform is offered: US East (N. Virginia), Canada (Central), Asia Pacific (Mumbai), Asia Pacific (Seoul), Asia Pacific (Sydney), Asia Pacific (Tokyo), Europe (Frankfurt), and Europe (London).
To learn more, see Modernizing .NET with AWS Transform in the AWS Transform User Guide.
AWS Lambda MicroVMs is now available in 7 additional AWS regions
AWS Lambda MicroVMs is now available in 7 additional AWS Regions: Asia Pacific (Osaka), Asia Pacific (Seoul), Canada (Central), Europe (London), Europe (Paris), South America (São Paulo), and US West (N. California). With this launch, you can use Lambda MicroVMs in a total of 21 AWS Regions.\n Lambda MicroVMs is a serverless compute primitive that provides VM-level isolation, near-instant launch and resume speeds, and state preservation for executing user or AI generated code. The service enables secure, stateful, and responsive code execution for applications including AI coding assistants, interactive development environments, and vulnerability scanners.
With today’s launch, developers can run Lambda MicroVMs closer to their end-users in 7 additional Regions, helping meet latency and data residency requirements. To get started, create a MicroVM image from your Dockerfile, then launch MicroVMs from that image. Give each user or job their own MicroVM with a dedicated HTTPS URL that supports popular connectivity protocols such as HTTP/2, gRPC, and WebSockets. You can use MicroVMs through the AWS Lambda console, AWS CloudFormation, AWS Cloud Development Kit, or with your preferred Agentic development tools using the Agent Toolkit for AWS.
To learn more, visit the AWS Lambda MicroVMs developer guide. For information on pricing, see Lambda MicroVMs pricing. Visit the AWS Capabilities by Region page for the latest region availability.
Introducing Amazon EC2 T9g instances based on Graviton5
AWS is announcing the general availability of new low-cost burstable Amazon EC2 T9g instances, which offer the best price performance among general-purpose EC2 instances. These instances are powered by AWS Graviton5 processors and built on the latest sixth generation AWS Nitro cards. T9g instances deliver up to 45% better price performance compared to previous generation T4g instances, and offer four sizes: nano, micro, small, and medium. These instances are designed for workloads requiring low to moderate CPU utilization, including data processing, login gateways, small databases, microservices, low-traffic websites, freemium services, staging and development, and short-duration compute tasks such as batch processing, event-driven functions, and CI/CD pipelines.\n T9g instances deliver up to 2.3x the compute performance, 1.5x higher network bandwidth, and 2.8x higher EBS bandwidth compared to T4g instances. They are built on the sixth-generation AWS Nitro System and feature the Nitro Isolation Engine, harnessing formal verification to provide mathematical assurance that customer workloads are isolated from each other and AWS operators, pioneering a new standard for mathematically proven cloud security.
T9g instances use the same CPU credit system as T4g with Standard and Unlimited modes, so existing T customers can upgrade to T9g and immediately benefit from better price performance, enabling them to lower their TCO. For customers new to AWS or migrating from on-premises, T9g instances provide one of the most cost-effective entry points to run a variety of lightweight workloads.
T9g.micro and T9g.small are available under the AWS Free Tier. T9g instances are available in US East (N. Virginia, Ohio), US West (N. California, Oregon), Canada (Central), South America (São Paulo), Europe (Frankfurt, Spain), and Asia Pacific (Malaysia, Singapore, Sydney, Tokyo) regions. T9g instances are available to purchase via Savings Plans, On-Demand instances, and Spot instances. To learn more, visit the Amazon EC2 T9g instances page.
Amazon MSK now supports in-place upgrade from Standard brokers to Express brokers
You can now upgrade existing Amazon MSK Provisioned clusters from Standard brokers to Express brokers in place, with fully managed data migration and no application changes. Express brokers are designed to deliver up to 3x more throughput per broker, scale up to 20x faster, and recover up to 90% faster compared to Standard brokers. With in-place upgrade, you can get the performance and operational benefits of Express brokers in hours, without needing to create a new cluster or having to migrate your data and applications to it. The upgrade keeps cluster endpoints unchanged, eliminating application cutovers, and preserves consumer offsets, so stateful applications such as Apache Flink and Kafka Streams continue without reprocessing or duplicates.\n Express brokers make Apache Kafka simpler to manage and more cost-effective to run at scale, with pay-as-you-go storage that scales automatically and built-in best practice defaults for higher resilience. Previously, upgrading to Express brokers meant creating a new cluster and migrating workloads to it, which required careful planning and cross-team coordination. With this launch, you can upgrade your existing Standard brokers to Express brokers in-place, without interrupting your applications, while maintaining complete control over the upgrade process. The upgrade adds Express brokers to your cluster and moves your topics to them, either automatically or in batches you control. Your Standard brokers stay in the cluster until you choose to decommission them to complete the upgrade, so you can validate your applications and roll back if needed.
In-place upgrade is available for MSK Provisioned clusters running Standard brokers on Apache Kafka version 3.9 with KRaft, in all AWS Regions where Express brokers are available. You can upgrade to Express brokers on M7g or M8g instances. Express brokers on M8g deliver up to 30% cost savings compared to Express brokers on M7g. There’s no additional charge for the upgrade itself. You pay for both Standard and Express brokers while the upgrade is in progress. To get started, visit the Amazon MSK Developer Guide and the Migration to Express brokers product page.
AWS launches standardized cost and rate columns in CUR 2.0 and Cost Explorer
AWS Data Exports for Cost and Usage Report (CUR) 2.0 now includes standardized cost and rate columns that represent cost consistently across AWS purchase options, including On Demand, Savings Plans, Reserved Instances, EC2 Spot, and EC2 Capacity Blocks for ML. These 14 columns are added when you enable Include Standardized cost and rate columns while setting up a new CUR 2.0 export or editing an existing CUR 2.0 and these columns have the prefix cost_analytics_. With these new columns, you can quantify the savings you realize from commitments, discounts, and tiered pricing, and analyze amortized cost the same way for every pricing model with multiple payment options, directly in CUR 2.0.\n Previously, calculating savings required combining CUR with Savings Plans rates and public price lists, and amortized cost used different columns for each pricing model. Now each line item carries billed, effective, list, contracted, public on-demand cost, the public equivalent of billed and effective cost before discounts, and a rate for each, so you can measure savings and amortized cost without building your own calculations. To add the columns, select Include standardized cost and rate columns when you create or edit a CUR 2.0 export. In Cost Explorer, choose the new cost types under Advanced options to see the same costs.
Standardized cost and rate columns are available at no additional charge in all AWS Regions except the AWS GovCloud (US) Regions and the China Regions. To learn more, see the CUR 2.0 data dictionary and the user guide for Cost Explorer.
AgentCore runtime instances now support Amazon EC2 G7 instances
Amazon Bedrock AgentCore runtime instances now support Amazon EC2 G7 instances, powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs. Runtime instances run your agents on Amazon EC2 instances while AgentCore handles provisioning, patching, scaling, and lifecycle management. G7 instances give agents that do 3D rendering or model inference another GPU option, alongside families such as G6e and G7e.\n To use G7, add G7 instance types, from g7.2xlarge to g7.48xlarge, to the allowed instance types of your capacity provider. AgentCore installs the NVIDIA drivers on the instance, so standard container images run GPU workloads without bundling drivers. For example, an agent can render 3D scenes or run a vision model on its own GPU, in sessions of up to 14 days.
G7 instances on runtime instances are available in select AWS Regions: US East (N. Virginia), US East (Ohio), US West (Oregon), Europe (Frankfurt), Europe (London), Europe (Stockholm), Asia Pacific (Mumbai), and Asia Pacific (Sydney). You are charged for managing the provisioned compute, in addition to your Amazon EC2 costs.
To get started, see the following resources:
Use GPU instance types in the Amazon Bedrock AgentCore Developer Guide
AgentCore pricing
Announcing the general availability of a new AWS Local Zone in Bogotá, Colombia
Today, AWS announces the general availability of a new Local Zone in Bogotá, Colombia, bringing AWS infrastructure closer to end users for applications that require single-digit millisecond latency. You can use the Bogotá Local Zone for workloads such as real-time gaming, live video streaming, medical imaging, and financial transaction processing.\n AWS Local Zones are infrastructure deployments that place compute, storage, networking, and other select AWS services closer to metropolitan areas, with the same APIs and tools as AWS Regions. You pay only for what you use, with no minimum commitment. Pricing options include On-Demand, Savings Plans, and Spot Instances. AWS Local Zones are available in more than 30 locations worldwide.
The AWS Local Zone in Bogotá supports Amazon Elastic Compute Cloud (Amazon EC2) with C7i, M7i, and R7i instances, Amazon S3 with the One Zone-Infrequent Access storage class, Amazon EBS with local snapshots and volume types gp3, gp2, and io1, Amazon Elastic Container Service (Amazon ECS), Amazon Elastic Kubernetes Service (Amazon EKS), Amazon Virtual Private Cloud (Amazon VPC), AWS Direct Connect, and Application Load Balancer.
To get started, enable the AWS Local Zone in Bogotá (us-east-1-bog-1a) from the Regions and Zones tab in the AWS Global View or by using the ModifyAvailabilityZoneGroup API. For pricing information, visit the AWS Local Zones pricing page. To learn more, visit the AWS Local Zones overview page.
AWS Continuum for Penetration Testing now estimates task hours before you run a test
AWS Continuum for Penetration Testing (formerly AWS Security Agent) now lets you estimate task hours before running a penetration test, so developers and security teams can see how many billable task hours a test is likely to need and choose a maximum task-hours limit with confidence. Continuum for Penetration Testing runs on-demand, AI-driven penetration tests against your web applications and APIs. Previously, the only way to know how long a test would take was to run it, making it difficult to plan budgets and set appropriate usage limits.\nWith task-hour estimates, you open a penetration test, choose the Pentest cost estimate tab, and request an estimate. AWS Continuum runs a short preliminary analysis of your application’s scope. It then draws on expertise from past penetration tests to estimate the task hours needed and shows the task hours that most similar applications finish within, and the higher range a few applications require. Requesting an estimate is free, and the estimate is a guide, not a quote or a limit. Actual task hours can be higher or lower, and only the maximum task-hours limit caps usage. Task-hour estimates are available in all AWS Regions where AWS Continuum for Penetration Testing is available. To learn more, see the documentation. To get started, visit the AWS Continuum product page.
Amazon Transcribe now available in AWS European Sovereign Cloud
Amazon Transcribe is now available in the AWS European Sovereign Cloud (Germany) Region. Customers who need to keep data and operations within the EU for regulatory compliance and data sovereignty requirements can now use automatic speech recognition (ASR) in the Region to turn speech into text.\n Amazon Transcribe offers two ways to transcribe audio in this Region. Batch transcription processes media files stored in Amazon S3, and streaming transcription returns text in real time from live audio over HTTP/2 or WebSocket. Both support custom vocabularies, so you can improve recognition of domain-specific terms such as product names, industry jargon, and proper nouns.
To learn more, see Getting started with Amazon Transcribe, Supported languages and language-specific features, and AWS European Sovereign Cloud pricing.
Amazon VPC Lattice now supports CloudWatch metrics for VPC resources
Amazon VPC Lattice now publishes Amazon CloudWatch metrics for VPC resources. Resource owners — those who share resources via VPC Lattice — and resource consumers — those who access those shared resources — can now see TCP traffic to resources such as databases, clusters, and on-premises endpoints. Resource owners get metrics for resource gateways and resource configurations. Consumers get metrics for resource endpoints, service network endpoints, and service network VPC associations. Previously, resource access logs were the primary way to see this traffic, but they required additional processing and provided limited real-time visibility. Now you can monitor your layer-4 resources directly in CloudWatch.\n With these metrics, you can proactively catch connectivity issues and optimize resource capacity. Capacity planning metrics — including bytes processed, new flows, and active flows — let resource owners monitor traffic per resource gateway and per resource configuration to track load on shared resources. Reliability metrics, such as dropped packets, help you detect problems early. Resource gateways also report port allocation errors, which show when a resource gateway needs more IPv4 addresses per network interface. Consumers can monitor traffic per resource endpoint, service network endpoint, or service network VPC association. They can set CloudWatch alarms for when new flows drop to zero or dropped packets increase, catching an unreachable resource before their applications do. Both owners and consumers can break metrics down by Availability Zone to isolate zonal issues or verify connections after configuration changes.
CloudWatch metrics for VPC resources are available in all AWS Regions where Amazon VPC Lattice is available. For the list of Regions, see the Amazon VPC Lattice FAQs.
To learn more, see CloudWatch metrics for Amazon VPC Lattice in the Amazon VPC Lattice User Guide, or visit the Amazon VPC Lattice product page.
AWS DevOps Agent adds community and partner tools repository
AWS DevOps Agent now gives you ways to share and leverage skills, MCP servers, connectors, and custom agents that enhance the capabilities of your AWS DevOps Agent instances.\nThe AWS DevOps Agent Tools open source repository is a centralized, AWS-curated collection of tools that extend AWS DevOps Agent capabilities. Customers can import ready-to-use, quality-controlled tools directly into their DevOps Agent configurations to improve investigation quality and expand operational context. A growing community of AWS DevOps Agent users and partners, and AWS service and support teams, have contributed a growing collection of tools to the community, including several partner-managed connectors that simplify connecting AWS DevOps Agent to third party MCP servers. Partners providing connectors include: Bluebox, Bronto, Celonis, Clickhouse Observability, Coralogix, Cribl, Elastic, Halo, Honeycomb, LaunchDarkly, Neo4j, Sumo Logic, and Wiz. To get started, browse the skills and connector catalogs from within your AWS DevOps Agent instance, or add and share your own tools by creating a PR within the AWS DevOps Agent Tools repository. See all the latest AWS DevOps Agent features on the release history page. For the list of AWS Regions where AWS DevOps Agent is available, see the supported Regions table.
AWS DevOps Agent adds new chat capabilities
AWS DevOps Agent now extends its chat capabilities across MCP clients like Kiro and Claude Code, as well as Slack and Microsoft Teams channels. With Model Context Protocol (MCP) support, AWS DevOps Agent can now take directed chat actions including support for GitHub, GitLab, Azure DevOps, ServiceNow, and PagerDuty MCP tools, giving developers and operations teams powerful workflow automation directly from their preferred chat interface.\nThis launch also adds auto-creation of per-incident Slack code channels and bi-directional chat within Microsoft Teams. AWS DevOps Agent has access to a compute sandbox that it can use to execute scripts and access the network to work more efficiently, and users can now rename and archive chats across all supported platforms. See all the latest AWS DevOps Agent features on the release history page. For the list of AWS Regions where AWS DevOps Agent is available, see the supported Regions table.
AWS DevOps Agent reduces investigation latency
When an incident is occurring, mean time to recover needs to be as fast as possible.\nWith this launch, incident investigations performed by AWS DevOps Agent complete faster. Investigations now also include a real-time summary that is updated with the current state of an investigation, so initial findings and observations surface quickly, before a final root cause and mitigation plan are created. For customers using AWS DevOps Agent’s built in Slack integration, these summaries are also automatically pushed to the relevant Slack channel. See all the latest AWS DevOps Agent features on the release history page. For the list of AWS Regions where AWS DevOps Agent is available, see the supported Regions table.
AWS Blogs
AWS Japan Blog (Japanese)
- Introducing Kiro Crew’s Conductor Mode
- 9th AWS Japan Generated AI Frontier Meetup ~A Place for Learning and Connection~ [Event Report]
- Until BTM Co., Ltd. rearchitected system investigation agents with Amazon Bedrock AgentCore
- AWS Lambda now supports GPU instances for event functions on Lambda Managed Instances
- 4 Continents, 4 New AWS Heroes: Closing 2026
- Koei Tecmo Games “Nobunaga’s Ambition Goes Out” has migrated its API server to AWS Graviton 4, achieving significant cost savings and improved response performance
AWS Architecture Blog
- How Bee implemented Zero Operator Access for personal AI on AWS
- AI coding agents on Amazon EC2 Mac: How Cognition scales Devin for macOS
AWS Cloud Financial Management
AWS Cloud Operations Blog
AWS Contact Center
- Modalities and Live Sync: give agents the screen in Amazon Connect Customer
- Agentic voices in Amazon Connect Customer: one voice from start to finish
- The Blind Spot in Contact Center Observability
AWS DevOps & Developer Productivity Blog
- Announcing custom execution roles for AWS CloudFormation StackSets
- Track AWS CloudFormation StackSets deployments with progress visibility
AWS for Industries
- The Trade Desk’s Cloud Journey: AI-Assisted Optimization to AI Agents
- How publishers can prepare their content for AI monetization on AWS
- Introducing AWS RTB Fabric Modules available through AWS Marketplace: AdTech partner intelligence, inline in the bid path
Artificial Intelligence
- How Standard Bank Group automates AI risk assessment for generative AI
- Announcing Amazon Textract Insights: Verifiable generative AI for documents
- Introducing enhanced infrastructure observability in Amazon SageMaker HyperPod
- Amazon SageMaker Inference: 2026 launches in review, part 2
- Use Kimi Code CLI with models on Amazon Bedrock