SageMaker MLflow now enables customers to encrypt their data using customer-managed keys (CMK) through AWS Key Management Service (KMS).
This enhancement allows organizations with strict security and compliance requirements to manage their own encryption keys. With customer-managed keys, you gain enhanced security control and comprehensive audit capabilities through AWS CloudTrail integration. You can encrypt your data with your own KMS keys, trace all data access for security auditing.
Customer-managed keys must be created in the same AWS account and region as your MLflow App, and only symmetric AWS KMS keys are supported.
SageMaker MLflow now enables customers to encrypt their data using customer-managed keys (CMK) through AWS Key Management Service (KMS).
This enhancement allows organizations with strict security and compliance requirements to manage their own encryption keys. With customer-managed keys, you gain enhanced security control and comprehensive audit capabilities through AWS CloudTrail integration. You can encrypt your data with your own KMS keys, trace all data access for security auditing.
Customer-managed keys must be created in the same AWS account and region as your MLflow App, and only symmetric AWS KMS keys are supported.
This feature is generally available in all AWS Regions where MLflow App is available. To learn more, visit the SageMaker MLflow detail page.
Reflexiones sobre el quinto aniversario de Windows 365
Windows 365 cumple cinco años y los PCs en la nube son reconocidos por permitir espacios de trabajo flexibles y confiables a gran escala
Por: Stefan Kinnestrand, vicepresidente de marketing comercial de Windows.
Hace cinco años, presentamos Windows 365 y creamos una nueva categoría de cómputo personal: el Cloud PC, que ofrece un escritorio Windows familiar que funciona en Microsoft Cloud, seguro por diseño, sencillo de gestionar y que permite trabajar desde cualquier dispositivo. En el quinto aniversario de Windows 365, reflexionamos sobre lo que han logrado nuestros clientes, para destacar los avances clave que han dado reconocimiento en la industria y compartir ideas sobre dónde deberían centrarse los líderes a continuación. La manera en que se realiza el trabajo ha cambiado mucho con la creciente adopción de IA y agentes, pero la necesidad de espacios de trabajo seguros, gestionados y fiables todavía existe.
El impacto medible de los PC en la nube con Windows 365
Hoy en día, decenas de miles de organizaciones de diferentes sectores y de todo el mundo utilizan PCs con Windows 365 Cloud. Los comentarios de los clientes cuentan una historia constante de cómo se desbloquean beneficios significativos, incluida la reducción de costes, menor latencia de red, menos tickets de soporte y una incorporación más rápida de empleados.
Miren lo que dicen los clientes sobre Windows 365. Más información en aka.ms/Windows365Stories.
En los últimos cinco años, hemos evolucionado Windows 365 de manera significativa e introducido nuevas ofertas para satisfacer las diversas necesidades de nuestros clientes:
Windows 365 Flex ofrece PCs en la nube para uso a tiempo parcial, por turnos u ocasional
Windows 365 Reserve equipa a los empleados con ordenadores en la nube temporales si sus dispositivos físicos no están disponibles
Windows 365 Link es el dispositivo sencillo, seguro y diseñado en específico para Windows 365, útil de manera específica para espacios de trabajo compartidos
Windows 365 para Agentes proporciona un entorno de ejecución seguro para que los agentes de IA puedan ejecutarse.
Estas soluciones facilitan que más clientes adopten experiencias de Windows basadas en la nube de formas que se adapten a las necesidades de su lugar de trabajo. Nuestro compromiso continuo de proporcionar espacios de trabajo seguros, gestionados y fiables para fuerzas laborales humanas y agentes a gran escala ha valido reconocimiento en la industria.
Microsoft reconocido como líder en el Magic Quadrant 2026 de Gartner® para Desktop como Servicio
Nos sentimos honrados de compartir que Microsoft ha sido nombrado Líder durante 4 años consecutivos en el Cuadrante Mágico de Gartner® para Escritorio como Servicio y ha ocupado el mejor puesto en Capacidad de Ejecución. Gartner define el escritorio como un servicio (DaaS, por sus siglas en inglés) como la provisión de escritorios virtuales por parte de proveedores de servicios en la nube pública u otros servicios. Lean una copia gratuita del informe para ver por qué Microsoft fue nombrado Líder.
Este gráfico fue publicado por Gartner, Inc. como parte de un documento de investigación más amplio y debe evaluarse en el contexto del documento completo disponible en aka.ms/DaaSMQ2026. Nota: Gartner no respalda ninguna empresa, proveedor, producto o servicio que aparezca en sus publicaciones, y no aconseja a los usuarios de tecnología seleccionar sólo a aquellos proveedores con las calificaciones u otra designación más alta. Las publicaciones de Gartner contienen las opiniones de la organización de análisis empresariales y tecnológicos de Gartner y no deben interpretarse como hechos reales. Gartner renuncia a todas las garantías, expresas o implícitas, respecto a esta publicación, incluida cualquier garantía de comerciabilidad o idoneidad para un propósito particular. Gartner, Magic Quadrant for Desktop as a Service, 5 de agosto de 2026, por Stuart Downes, Todd Larivee, Sunil Kumar. Gartner y Magic Quadrant son marcas registradas de Gartner, Inc. y/o sus filiales.
Microsoft reconocido como líder en la evaluación de proveedores IDC MarketScape: Worldwide Desktop as a Service 2026
El modelo de análisis de proveedores IDC MarketScape está diseñado para ofrecer una visión general de la aptitud competitiva de los proveedores de TIC en un mercado determinado. La metodología de investigación utiliza una rigurosa puntuación basada en criterios tanto cualitativos como cuantitativos, que da como resultado una única ilustración gráfica de la posición de cada proveedor en un mercado determinado. La puntuación de Capacidades mide el producto del proveedor, la salida al mercado y la ejecución empresarial a corto plazo. La puntuación de Estrategia mide la alineación de las estrategias de los proveedores con los requisitos del cliente en un periodo de 3 a 5 años. La cuota de mercado de los vendedores se representa por el tamaño de los iconos.
«Microsoft ofrece una pila DaaS integrada por completo donde la gestión de terminales, identidad, seguridad y cumplimiento actúan como componentes nativos preconstruidos.» Además, «Microsoft destaca como el único proveedor que comercializa el espacio de trabajo agente como una oferta distinta de compra junto al escritorio gestionado estándar. Esto proporciona una base sólida de producción para los líderes de TI cuyas estrategias DaaS deben acomodar el despliegue de agentes de IA durante los próximos tres años.»
– IDC MarketScape: Evaluación Mundial de Escritorios como Servicio 2026 (Doc #US54118526, julio de 2026)
¿Qué hay de nuevo y qué viene con Windows 365?
Cinco años después, los PC en la nube permiten un trabajo seguro y flexible no solo para los trabajadores de la información, sino también para la primera línea, desarrolladores y agentes emergentes de IA. Estamos entusiasmados de compartir nuestros desarrollos más recientes y hacia dónde se dirigen las cosas a continuación. Miren esta demo de Microsoft Mechanics para echar un vistazo más de cerca y lean a continuación para saber más:
Empoderar a los desarrolladores
Para los desarrolladores, hicimos varios anuncios en Microsoft Build 2026, incluidos PCs en la Nube listos‑para‑programar preconfigurados con herramientas de desarrollo comunes, que ahora están en vista previa. También anunciamos la disponibilidad general de una nueva opción Windows 365 GPU Select junto con PCs Windows 365 Cloud de 32vCPUpara cargas de trabajo intensivas en gráficos y computación como desarrollo de software, modelado de datos, simulaciones e IA/ML. Algunos modelos de lenguaje (LM, por sus siglas en inglés) pueden ejecutarse directo en PCs en la nube, para permitir a los desarrolladores que crean aplicaciones impulsadas por IA utilizar la computación en la nube para construir e iterar de manera eficaz sin incurrir en costes de tokens. Los PCs en la nube siempre disponibles también permiten que compilaciones de larga duración, bucles de prueba, entornos de desarrollo y flujos de trabajo asistidos por IA sean persistentes más allá de los límites de un dispositivo físico, para que el trabajo pueda continuar en un PC en la nube incluso cuando el dispositivo físico entra en suspensión.
Habilitación de agentes de IA
Para los que hacen la producción de agentes, hicimos Windows 365 para Agentes disponible a nivel general a principios de este año. Windows 365 para Agentes ofrece una colección de sistemas conectados a Microsoft Entra, gestionados por Microsoft Intune y aplicados por políticas de PCs en la nube para que los agentes ejecuten flujos de trabajo en varios pasos a través del software, incluido abrir aplicaciones, navegar por interfaces, introducir entradas y procesar datos. Empresas como Login VSI, Simular, y WorkspaceDNA ya han comenzado a desarrollar agentes de IA que pueden ejecutarse en Windows 365 for Agents. Y en Microsoft, usamos Windows 365 para Agentes para impulsar experiencias agénticas, como escenarios de uso de ordenadores en Researcher.
De cara al futuro, seguimos con la introducción de nuevas capacidades para permitir el uso seguro de agentes de IA de nivel empresarial. Ahora en vista previa, Windows 365 for Agents permite acceder a aplicaciones locales por medio de la identidad de los agentes, lo que permite a las organizaciones extender la automatización a través de aplicaciones empresariales heredadas sin modernizar las aplicaciones existentes. También en la vista previa, introdujimos soporte para vincular los pools de agentes Cloud para PC de Windows 365 for Agents con un grupo de Microsoft Entra o con un perfil de preparación de dispositivos Microsoft Intune Autopilot. Esto permite a los administradores de TI gestionar los PCs en la nube para agentes a gran escala a través de la aplicación de políticas de seguridad y cumplimiento, para desplegar las aplicaciones necesarias, mantener configuraciones consistentes entre los pools de agentes y utilizar flujos de trabajo existentes en Intune. Además, con el soporte de Windows 365 para Microsoft Execution Containers (MXC) que estará disponible en los próximos meses, los clientes pueden ejecutar de manera segura cargas de trabajo de agentes compatibles, desde agentes de codificación, como la CLI de GitHub Copilot, hasta agentes autónomos como OpenClaw, en PCs en la Nube gestionados y siempre disponibles, por medio deaislamiento de procesos y sesiones con controles familiares de seguridad y gobernanza empresarial.
Fortalecimiento de la seguridad
Nuestro compromiso con la seguridad sigue tan importante como siempre. Los PC en la nube permiten a las organizaciones reducir su superficie de ataque a través de mantener los datos seguros en Microsoft Cloud, no en puntos finales dispersos. A principios de este año, introdujimos redireccionamientos basados en el contexto a través del contexto de autenticación de Microsoft Entra, lo que permite un control más preciso sobre cómo los usuarios acceden a los PCs en la nube. Reforzamos aún más las defensas frente a amenazas de terminales a través de la protección de datos sensibles en pantalla. Con las capacidades de protección de salida que estarán disponibles a nivel general en septiembre, el contenido mostrado en un PC en la nube puede mantenerse protegido de los screen scrapers y otras amenazas en las terminales.
Optimización de la gestión
También facilitamos a los administradores de TI optimizar aún más la gestión de PC en la nube a través del aprovechamiento de la IA. Con la nueva plataforma de monitorización e informes de Windows 365 que estará disponible en septiembre y un nuevo agente para administradores de Windows 365 ya disponible en vista previa limitada, los administradores de TI pueden monitorizar con mayor facilidad el estado del PC en la nube en Microsoft Intune, investigar problemas, identificar causas raíz probables y acelerar la resolución de problemas mediante una experiencia conversacional guiada que agiliza el análisis de señales y conocimientos operativos en todo el PC en la nube medio ambiente. Completen este formulario o contacten con su equipo de cuentas de Microsoft para expresar interés en participar en la vista previa del agente de administrador de Windows 365.
Plataforma de monitorización e informes de Windows 365 (disponible a nivel general en septiembre) y un nuevo agente para administradores de Windows 365 (vista previa limitada).
Colaborar en el próximo capítulo
De cara al futuro, seguimos comprometidos a evolucionar para satisfacer las diversas necesidades de los clientes, a través de la habilitación de nuevas maneras de trabajar de forma segura, impulsar la productividad humana y agéntica y asegurar la continuidad del negocio. Escuchen a Judson Althoff, CEO de Microsoft Commercial Business, sobre cómo una base preparada para IA que sirve a empleados y agentes puede ayudar a cada empresa a ofrecer la productividad, flexibilidad y seguridad que su plantilla necesita para prosperar.
Estamos agradecidos por los clientes y socios que han dado forma a este camino con nosotros, y estamos entusiasmados por construir juntos el próximo capítulo de los PC en la nube.
¿Listos para empezar?
Contacten con ventas para hablar sobre casos de uso de Windows 365 en trabajo de información, trabajo de primera línea y desarrollo de software y evaluar planes de PC en la nube.
¿Buscan un entorno de ejecución seguro para agentes emergentes? Prueben 50 horas gratuitas de Windows 365 para Agents Cloud PC con Microsoft Copilot Studio.
¿Buscan una solución de continuidad de negocio? Consideren Windows 365 Reserve para equipar a los empleados con PCs en la nube si sus dispositivos principales dejan de estar disponibles.
¿Evalúan una actualización de PC? Consideren Windows 365 Link —el dispositivo sencillo, seguro y diseñado en específico para Windows 365— como una alternativa rentable a los escritorios tradicionales.
Amazon Aurora PostgreSQL-Compatible Edition now supports PostgreSQL versions 18.4, 17.10, 16.14, 15.18, and 14.23 which include bug fixes from the PostgreSQL community and Aurora-specific enhancements. We recommend upgrading to the latest minor versions to address known Common Vulnerabilities and Exposures (CVEs) and benefit from these improvements, as detailed in the release notes.
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 multiple upgrades in phases, validating on lower-priority environments before upgrading your most critical ones. For more information, see Upgrading Amazon Aurora PostgreSQL DB clusters.
Amazon Aurora is designed for high performance and availability at global scale with full PostgreSQL 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 our getting started page.
Amazon Aurora PostgreSQL-Compatible Edition now supports PostgreSQL versions 18.4, 17.10, 16.14, 15.18, and 14.23 which include bug fixes from the PostgreSQL community and Aurora-specific enhancements. We recommend upgrading to the latest minor versions to address known Common Vulnerabilities and Exposures (CVEs) and benefit from these improvements, as detailed in the release notes.
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 multiple upgrades in phases, validating on lower-priority environments before upgrading your most critical ones. For more information, see Upgrading Amazon Aurora PostgreSQL DB clusters.
Amazon Aurora is designed for high performance and availability at global scale with full PostgreSQL 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 our getting started page.
Amazon SageMaker HyperPod now enhances support for Ray with built-in observability, resilient training, accelerated inference and managed development environments. Ray is a popular open-source framework for scaling AI workloads on a unified compute layer, from data processing and distributed training to reinforcement learning and model serving. Running Ray on Kubernetes at production scale can be an operational burden: job hangs, low GPU utilization from static team allocations, and multi-step observability setup. Also, lack of interactive development environment means every code change needs another job submission and familiarity with kubectl.
HyperPod now brings easier development, resilient training, and accelerated inference to Ray. Data scientists create, edit, monitor, and delete Ray clusters from a web-based interface in Amazon SageMaker Studio, then attach JupyterLab, Code Editor, or a local IDE to a running Ray cluster and iterate interactively against cluster-scale compute. A multi-node Ray cluster behaves like a local development environment, so you test each change immediately, without waiting for a new job to queue and start. For Observability, HyperPod provisions Grafana dashboards with metrics in Amazon Managed Service for Prometheus and allows one-click access to the Ray Dashboard through a secure browser link, giving you visibility into your workloads from the first run. For training at scale, HyperPod node auto recovery and hung job detection handle GPU faults, job hangs, loss spikes, and degraded throughput. Tiered checkpointing restores state from cluster memory to maximize goodput, and task governance improves compute utilization through quotas, priorities, and preemption. Together, these keep your long training runs progressing through failures and maximize the useful work done per GPU-hour. For inference with Ray Serve, a tiered KV cache reuses cached prefixes to reduce time to first token, and you can deploy Amazon SageMaker JumpStart models directly.
Open-source Ray code runs unchanged and you can either adopt the purpose-built experience in SageMaker Studio or take individual capabilities to integrate into your own ML platform.
Ray support is available for HyperPod clusters orchestrated by Amazon EKS, in AWS Regions where SageMaker HyperPod is supported. To learn more, see the SageMaker HyperPod documentation, and explore the interactive demo.
Amazon SageMaker HyperPod now enhances support for Ray with built-in observability, resilient training, accelerated inference and managed development environments. Ray is a popular open-source framework for scaling AI workloads on a unified compute layer, from data processing and distributed training to reinforcement learning and model serving. Running Ray on Kubernetes at production scale can be an operational burden: job hangs, low GPU utilization from static team allocations, and multi-step observability setup. Also, lack of interactive development environment means every code change needs another job submission and familiarity with kubectl.
HyperPod now brings easier development, resilient training, and accelerated inference to Ray. Data scientists create, edit, monitor, and delete Ray clusters from a web-based interface in Amazon SageMaker Studio, then attach JupyterLab, Code Editor, or a local IDE to a running Ray cluster and iterate interactively against cluster-scale compute. A multi-node Ray cluster behaves like a local development environment, so you test each change immediately, without waiting for a new job to queue and start. For Observability, HyperPod provisions Grafana dashboards with metrics in Amazon Managed Service for Prometheus and allows one-click access to the Ray Dashboard through a secure browser link, giving you visibility into your workloads from the first run. For training at scale, HyperPod node auto recovery and hung job detection handle GPU faults, job hangs, loss spikes, and degraded throughput. Tiered checkpointing restores state from cluster memory to maximize goodput, and task governance improves compute utilization through quotas, priorities, and preemption. Together, these keep your long training runs progressing through failures and maximize the useful work done per GPU-hour. For inference with Ray Serve, a tiered KV cache reuses cached prefixes to reduce time to first token, and you can deploy Amazon SageMaker JumpStart models directly.
Open-source Ray code runs unchanged and you can either adopt the purpose-built experience in SageMaker Studio or take individual capabilities to integrate into your own ML platform.
Ray support is available for HyperPod clusters orchestrated by Amazon EKS, in AWS Regions where SageMaker HyperPod is supported. To learn more, see the SageMaker HyperPod documentation, and explore the interactive demo.
AWS ParallelCluster 3.16 is now generally available with a new on-node diagnostics tool, cluster stability improvements, and an updated HPC and AI/ML software stack.
pcluster-diag is a diagnostics tool built into the ParallelCluster AMIs that lets you run diagnostic checks on any cluster node with a single command, and get a structured report that makes it easier to identify issues. This release also hardens the cluster lifecycle with more resilient cluster creation, updates, and image builds. The software stack is refreshed, with updated NVIDIA driver, CUDA, EFA installer, and Slurm versions. To get started with pcluster-diag, see Troubleshooting with pcluster-diag. For more details, review the AWS ParallelCluster 3.16.0 release notes.
AWS ParallelCluster is an open-source cluster management tool that makes it possible for R&D customers and IT administrators to operate high-performance computing (HPC) clusters on AWS. ParallelCluster is designed to automatically and securely provision cloud resources into elastically-scaling HPC clusters capable of running scientific and engineering workloads at scale on AWS. ParallelCluster is available at no additional charge in the AWS Regions listed here, and you pay only for the AWS resources needed to run your applications.
To learn more about launching HPC clusters on AWS, visit the ParallelCluster User Guide. To start using ParallelCluster, see the installation instructions for ParallelCluster UI and CLI.
AWS ParallelCluster 3.16 is now generally available with a new on-node diagnostics tool, cluster stability improvements, and an updated HPC and AI/ML software stack.
pcluster-diag is a diagnostics tool built into the ParallelCluster AMIs that lets you run diagnostic checks on any cluster node with a single command, and get a structured report that makes it easier to identify issues. This release also hardens the cluster lifecycle with more resilient cluster creation, updates, and image builds. The software stack is refreshed, with updated NVIDIA driver, CUDA, EFA installer, and Slurm versions. To get started with pcluster-diag, see Troubleshooting with pcluster-diag. For more details, review the AWS ParallelCluster 3.16.0 release notes.
AWS ParallelCluster is an open-source cluster management tool that makes it possible for R&D customers and IT administrators to operate high-performance computing (HPC) clusters on AWS. ParallelCluster is designed to automatically and securely provision cloud resources into elastically-scaling HPC clusters capable of running scientific and engineering workloads at scale on AWS. ParallelCluster is available at no additional charge in the AWS Regions listed here, and you pay only for the AWS resources needed to run your applications. To learn more about launching HPC clusters on AWS, visit the ParallelCluster User Guide. To start using ParallelCluster, see the installation instructions for ParallelCluster UI and CLI.
GPT-5.6 Terra and Luna are now generally available on Amazon Bedrock in AWS GovCloud (US-West) and AWS GovCloud (US-East), bringing the smartest family of models from OpenAI yet to Bedrock’s next-generation inference engine built for high-performance, security and reliability. GPT-5.6 sets a new standard for intelligence and efficiency, allowing you to solve harder problems in less time and with more intelligence per token. The two models span capability tiers from balanced performance (Terra) to fast, cost-efficient inference (Luna).
With GPT-5.6, you can build autonomous coding agents, run long-horizon genomics and biology analyses, and perform advanced cybersecurity research. Terra provides GPT-5.5-level performance at half the cost and Luna brings fast, affordable inference at the lowest price point. GPT-5.6 also supports prompt caching with explicit cache breakpoints, so repeated context across agentic workflows is billed at a 90% discount and doesn’t compound cost as you scale.
GPT-5.6 Terra and Luna support 1 million token context windows on Amazon Bedrock, enabling you to process full codebases, lengthy documents, and multi-turn agent histories in a single request. Models reason over broader context and return more accurate, coherent responses without chunking or information loss.
GPT-5.6 Terra and Luna are now generally available on Amazon Bedrock in AWS GovCloud (US-West) and AWS GovCloud (US-East), bringing the smartest family of models from OpenAI yet to Bedrock’s next-generation inference engine built for high-performance, security and reliability. GPT-5.6 sets a new standard for intelligence and efficiency, allowing you to solve harder problems in less time and with more intelligence per token. The two models span capability tiers from balanced performance (Terra) to fast, cost-efficient inference (Luna).
With GPT-5.6, you can build autonomous coding agents, run long-horizon genomics and biology analyses, and perform advanced cybersecurity research. Terra provides GPT-5.5-level performance at half the cost and Luna brings fast, affordable inference at the lowest price point. GPT-5.6 also supports prompt caching with explicit cache breakpoints, so repeated context across agentic workflows is billed at a 90% discount and doesn’t compound cost as you scale.
GPT-5.6 Terra and Luna support 1 million token context windows on Amazon Bedrock, enabling you to process full codebases, lengthy documents, and multi-turn agent histories in a single request. Models reason over broader context and return more accurate, coherent responses without chunking or information loss.
For regional availability, please see the Amazon Bedrock regional availability page. Get started with Terra and Luna using the Amazon Bedrock Console or the Responses API on the bedrock-mantle endpoint. To learn more, see the Amazon Bedrock documentation and read the launch blog post.
Today, OpenAI announced that they are lowering API prices for GPT-5.6 Sol. Following the recent Terra and Luna price reductions, Sol now costs $4 per million input tokens and $20 per million output tokens—20% lower input pricing and 33.3% lower output pricing. This promotional pricing is available at least through November 21, 2026.
Whether you’re building autonomous coding agents, running complex multi-step analyses, or performing advanced research workflows, the reduced pricing gives you more room to experiment and scale what’s already working. GPT-5.6 Sol delivers state-of-the-art results on agentic coding benchmarks, and the lower price point makes it more accessible for sustained, high-volume workloads.
For latest Regional availability of GPT-5.6 Sol, check the AWS Regions page. To learn more and view the pricing visit the Amazon Bedrock documentation on GPT-5.6 Sol.
Today, OpenAI announced that they are lowering API prices for GPT-5.6 Sol. Following the recent Terra and Luna price reductions, Sol now costs $4 per million input tokens and $20 per million output tokens—20% lower input pricing and 33.3% lower output pricing. This promotional pricing is available at least through November 21, 2026.
Whether you’re building autonomous coding agents, running complex multi-step analyses, or performing advanced research workflows, the reduced pricing gives you more room to experiment and scale what’s already working. GPT-5.6 Sol delivers state-of-the-art results on agentic coding benchmarks, and the lower price point makes it more accessible for sustained, high-volume workloads.
For latest Regional availability of GPT-5.6 Sol, check the AWS Regions page. To learn more and view the pricing visit the Amazon Bedrock documentation on GPT-5.6 Sol.
The Amazon Elastic Kubernetes Service (Amazon EKS) Capability for Argo CD now supports custom configuration through a standard argocd-cm ConfigMap in your cluster. This capability gives you a fully managed GitOps continuous delivery experience, and you can now tune it to fit how your teams work. You can define custom health checks for your Custom Resources, customize the Argo CD UI banner content, adjust how the capability watches and compares the resources it manages, and more. You configure these settings the same way you do in upstream Argo CD, and AWS applies them to your managed capability.
With this launch, cluster administrators now have more control over how Argo CD reports application health. By default, Argo CD has no built-in health logic for Custom Resources, so an Application can report as healthy while its resources are still provisioning, and sync waves can advance before those resources are ready. With a custom health check, you define this logic yourself. For example, a health check for a database resource can hold an Application at progressing until the database is ready. The capability also includes built-in health checks for AWS Controllers for Kubernetes (ACK) and kro (Kube Resource Orchestrator) resources, so these report accurate health with no additional configuration.
You can configure the EKS Capability for Argo CD in all AWS Regions where the capability is available. To learn more, see Amazon EKS and Configure Argo CD settings in the Amazon EKS User Guide.
The Amazon Elastic Kubernetes Service (Amazon EKS) Capability for Argo CD now supports custom configuration through a standard argocd-cm ConfigMap in your cluster. This capability gives you a fully managed GitOps continuous delivery experience, and you can now tune it to fit how your teams work. You can define custom health checks for your Custom Resources, customize the Argo CD UI banner content, adjust how the capability watches and compares the resources it manages, and more. You configure these settings the same way you do in upstream Argo CD, and AWS applies them to your managed capability.
With this launch, cluster administrators now have more control over how Argo CD reports application health. By default, Argo CD has no built-in health logic for Custom Resources, so an Application can report as healthy while its resources are still provisioning, and sync waves can advance before those resources are ready. With a custom health check, you define this logic yourself. For example, a health check for a database resource can hold an Application at progressing until the database is ready. The capability also includes built-in health checks for AWS Controllers for Kubernetes (ACK) and kro (Kube Resource Orchestrator) resources, so these report accurate health with no additional configuration.
You can configure the EKS Capability for Argo CD in all AWS Regions where the capability is available. To learn more, see Amazon EKS and Configure Argo CD settings in the Amazon EKS User Guide.
Amazon Connect Customer now lets managers chat with their data in plain language and get back the answer, the evidence behind it, and the fix, in seconds. Managers have always had the data. What they haven’t had is the time to dig through dashboards, find what’s driving performance, and decide what to do next. Now Amazon Connect Customer does that work for them. It searches across more than 150 metrics spanning self-service, agent performance, and queue performance to find what matters, explain why, and recommend the best next step.
Managers can start broad and go deep in the same conversation. For example, a manager can ask which queues are the best candidates for automation, and Amazon Connect Customer reviews where handle time and after-contact work run highest, then returns a prioritized list with confidence scores and projected impact. What once required analysts, dashboards, and weeks of investigation now becomes a prioritized action plan in seconds.
Amazon Connect Customer now lets managers chat with their data in plain language and get back the answer, the evidence behind it, and the fix, in seconds. Managers have always had the data. What they haven’t had is the time to dig through dashboards, find what’s driving performance, and decide what to do next. Now Amazon Connect Customer does that work for them. It searches across more than 150 metrics spanning self-service, agent performance, and queue performance to find what matters, explain why, and recommend the best next step. Managers can start broad and go deep in the same conversation. For example, a manager can ask which queues are the best candidates for automation, and Amazon Connect Customer reviews where handle time and after-contact work run highest, then returns a prioritized list with confidence scores and projected impact. What once required analysts, dashboards, and weeks of investigation now becomes a prioritized action plan in seconds. This feature is available in all AWS Regions where Amazon Connect Customer AI Agents are supported. To learn more, visit our product documentation.
The AWS Deadline Cloud monitor now shows the progress, status, and health of your automatic file downlaods from jobs running in the cloud. Deadline Cloud is a fully managed service that helps teams run compute-intensive workloads in the cloud for visual effects, animation, product design, simulation, and gaming. The Deadline Cloud Monitor (DCM) desktop app provides customers with visibility into their render environments, jobs, resources, and costs. Now, customers can also use the monitor to confirm that automatically configured job outputs successfully downloaded to their destination drive.
With this update, the monitor app introduces a new Download status column at both the job and task level, showing download progress and confirming when all output files are available on your drive. An indicator displays how current that status is. If files are unavailable for any reason, the app surfaces clear guidance on next steps. This eliminates manual drive verification and helps teams confidently confirm output availability before downstream tasks begin, particularly valuable in large-scale render pipelines where manual file checking is impractical.
To learn more about AWS Deadline Cloud and the new automatic download status feature in the Deadline Cloud Monitor desktop app, visit https://aws.amazon.com/deadline-cloud/.
The AWS Deadline Cloud monitor now shows the progress, status, and health of your automatic file downlaods from jobs running in the cloud. Deadline Cloud is a fully managed service that helps teams run compute-intensive workloads in the cloud for visual effects, animation, product design, simulation, and gaming. The Deadline Cloud Monitor (DCM) desktop app provides customers with visibility into their render environments, jobs, resources, and costs. Now, customers can also use the monitor to confirm that automatically configured job outputs successfully downloaded to their destination drive.
With this update, the monitor app introduces a new Download status column at both the job and task level, showing download progress and confirming when all output files are available on your drive. An indicator displays how current that status is. If files are unavailable for any reason, the app surfaces clear guidance on next steps. This eliminates manual drive verification and helps teams confidently confirm output availability before downstream tasks begin, particularly valuable in large-scale render pipelines where manual file checking is impractical.
To learn more about AWS Deadline Cloud and the new automatic download status feature in the Deadline Cloud Monitor desktop app, visit https://aws.amazon.com/deadline-cloud/.
2026 de Gartner® para Desktop como Servicio
