Today, AWS announces the general availability of the Get Invoice PDF API, enabling customers to programmatically download AWS invoices via SDK calls.
Customers can retrieve individual invoice PDF artifacts by invoking API calls with AWS Invoice ID as input and receives pre-signed Amazon S3 URL for immediate download of AWS invoice and supplemental documents in PDF format. For bulk invoice retrieval, customers can first call the List Invoice Summaries API to get Invoice IDs for a specific billing period, then use the Invoice IDs as input to Get Invoice API to download each Invoice PDF artifact.
The Get Invoice PDF API is available in the US East (N. Virginia) Region. Customers from any commercial regions (except China Regions) can use the service. To get started with Get Invoice PDF API please visit the API Documentation.
Today, AWS announces the general availability of the Get Invoice PDF API, enabling customers to programmatically download AWS invoices via SDK calls. Customers can retrieve individual invoice PDF artifacts by invoking API calls with AWS Invoice ID as input and receives pre-signed Amazon S3 URL for immediate download of AWS invoice and supplemental documents in PDF format. For bulk invoice retrieval, customers can first call the List Invoice Summaries API to get Invoice IDs for a specific billing period, then use the Invoice IDs as input to Get Invoice API to download each Invoice PDF artifact. The Get Invoice PDF API is available in the US East (N. Virginia) Region. Customers from any commercial regions (except China Regions) can use the service. To get started with Get Invoice PDF API please visit the API Documentation.
Microsoft Ignite 2025: Copilot y agentes creados para impulsar la Empresa Frontera
Por: Jared Spataro, director de marketing, IA at Work, Microsoft.
El futuro del trabajo estará marcado por las Empresas Frontera (Frontier Firms), organizaciones lideradas por personas y gestionadas por agentes. Estas empresas han comenzado a transformar la manera en que se realiza el trabajo, al empoderar a cada empleado con un asistente de IA, lo que amplifica el impacto con el trabajo en equipo humano-agente y reinventa los procesos empresariales con los agentes. En Microsoft Ignite, hemos presentado nuevas capacidades en Microsoft 365 Copilot para ayudar a cada cliente a convertirse en Frontier.
Work IQ — la capa de inteligencia que permite a Microsoft 365 Copilot conocerlos, a ustedes, a su trabajo y a su empresa a la perfección.
Los agentes de Word, Excel y PowerPoint en el chat —junto con el Modo Agente en las aplicaciones de Office— les permiten trabajar de manera iterativa con Copilot para entregar documentos de Office, hojas de cálculo y presentaciones de alta calidad.
El Agent 365—el plano de control para agentes—les permite gestionar y asegurar a los agentes con confianza.
También nos complace anunciar la Iniciativa de IA de la Frontier Firm con el Instituto de Diseño de Datos Digitales de Harvard (D^3). La iniciativa tiene como objetivo investigar la colaboración entre humanos e IA y dotar a líderes globales de alto nivel con las habilidades necesarias para transformar sus organizaciones en Empresas Frontera. Barclays, BNY, Cigna Healthcare, Clifford Chance, DuPont, Eaton, Eli Lilly and Company, EY, GHD, Mastercard, Levi Strauss, Lumen, Mastercard y Nestlé están entre los primeros equipos.
Microsoft 365 Copilot con Work IQ: IA diseñada para el trabajo
Hacemos de Copilot un producto que les encantará y que no querrán vivir sin él. Ya, más del 90% de las empresas Fortune 500 utilizan Microsoft 365 Copilot. Innovamos a un ritmo acelerado, con el lanzamiento de más de 400 nuevas funciones en el último año y llevamos los últimos modelos a Copilot, ajustados para el trabajo y adaptados a las necesidades de su negocio—desde GPT-5 hasta el uso en ordenador (CUA, por sus siglas en inglés) y Sora 2. Y ampliamos la elección de modelos con la incorporación de modelos antrópicos para asegurar el modelo adecuado para cada trabajo.
Work IQ es la capa de inteligencia detrás de Microsoft 365 Copilot y sus agentes que ayuda a Copilot a entenderlos a ustedes, a su trabajo y a su empresa. Primero, son sus datos de trabajo—todo el conocimiento valioso en sus correos, archivos, reuniones y chats—que codifican cómo se realiza el trabajo. Segundo, la memoria—su estilo, preferencias, hábitos y flujos de trabajo—los patrones de trabajo y las relaciones únicas que les resultan especiales. No solo entiende su organigrama, sino también su cuadro de trabajo. En tercer lugar, la inferencia combina datos y memoria para establecer conexiones valiosas, desbloquear conocimientos y predecir la siguiente mejor acción, para ir mucho más allá de lo que pueden hacer los conectores. Por ejemplo, Copilot sugiere el agente adecuado para la tarea según su prompt e intención. Es un ciclo de retroalimentación impulsado por IA: Work IQ está integrado en las aplicaciones que usan cada día, como Word, Outlook y Teams, así que Copilot siempre aprende—y usa esa señal para ofrecer experiencias cada vez más personalizadas.
Con Work IQ para agentes personalizados, ahora pueden acceder a esta capa de inteligencia para crear agentes adaptados a sus flujos de trabajo únicos y necesidades empresariales. Esto permite un aterrizamiento seguro de agentes que respeta los permisos existentes, etiquetas de sensibilidad, controles de cumplimiento, auditoría, registro, monitorización y aplicación de políticas, y está disponible en Copilot Studio o mediante API para su uso en el desarrollo de agentes pro-code. Pueden usar Work IQ para agentes personalizados con una licencia Microsoft 365 Copilot o en facturación por consumo. Copilot y los agentes —impulsados por Work IQ— ponen todo el conocimiento de su empresa al servicio de ustedes y de su negocio.
Las actualizaciones de Microsoft 365 Copilot incluyen:
El mes pasado llevamos vibe working a Microsoft 365 Copilot con el Modo Agente en las aplicaciones de Office. Ahora pueden crear artefactos de Office de primer nivel, tanto si empiezan en el chat de Copilot como en las aplicaciones. Con el Modo Agente en Excel, los usuarios pueden ahora elegir entre modelos de razonamiento Anthropic y OpenAI. El Modo Agente en Word ya está disponible de forma general, y el Modo Agente en PowerPoint ya está disponible a través del programa Frontier. Y los nuevos Agentes de Word, Excel y PowerPoint les permiten crear documentos, hojas de cálculo y presentaciones de alta calidad en el chat de Copilot. Disponible a través del programa Frontier.
Ahora pueden hablar con Copilot como un compañero con voz en la app Microsoft 365 Copilot. Desde su móvil, pregunten a Copilot: «¿cuáles son mis prioridades principales para el día?» o «ponme al día con la reunión que me perdí» (disponible a nivel general en diciembre). Copilot en Outlook facilita el triaje de su bandeja de entrada y el calendario mientras están en movimiento, solo con su voz (en vista previa pública). También pueden leer y responder con rapidez a correos electrónicos con nuevas preguntas de un solo toque como «resume y responde», y programar reuniones, todo desde su teléfono (disponible a nivel general).
Copilot Chat es su chat seguro de IA para el trabajo, disponible para cualquier suscriptor de Microsoft 365 sin coste adicional. Y pronto, incluirá acceso al Modo Agente en el chat y las aplicaciones de Office y Copilot en Outlook entenderán su bandeja de entrada y su calendario. Llegará en vista previa a principios de 2026.
Sora 2 pone en marcha la creación de vídeo con IA de nueva generación. A través de Crear, ahora pueden crear, editar y compartir vídeos cortos para marketing o contenido en redes sociales. Disponible a través del Programa Frontier.
Edge for Business ya es de confianza entre los principales clientes durante la mayor parte de sus minutos de navegación, y con el Modo Copilot activado, se convierte en el primer navegador de IA empresarial seguro del mundo. Los flujos de trabajo de varios pasos están habilitados con una licencia Microsoft 365 Copilot, y el razonamiento multi-pestañas llegará pronto a través del programa Frontier.
Hacemos realidad el trabajo en equipo entre humanos y agentes. El Modo Teams para Microsoft 365 Copilot convierte sus chats 1:1 en Copilot en chats grupales en Teams (en vista previa pública). También introdujimos agentes centrados en la colaboración para Copilot—asignando a cada equipo, proyecto y encuentro con un compañero de IA. El agente facilitador en Teams dirige la agenda, toma notas, mantiene las reuniones en orden e incluso ayuda a gestionar las acciones de la reunión (disponibles a nivel general). Ahora, los agentes en los canales de Teams pueden trabajar con aplicaciones y agentes de terceros a través de servidores Model Context Protocol (MCP), como GitHub, Asana y Atlassian (Jira). Por ejemplo, un usuario puede preguntar al agente en un canal de Teams sobre bloqueadores y planes de mitigación para el próximo lanzamiento de un producto, y el agente extraerá riesgos directo de Jira para luego programar una reunión con el equipo (en vista previa pública).
Copilot y agentes transforman cada proceso empresarial
Queremos ayudar a cada cliente a transformar procesos de negocio con agentes, y Copilot es su ventana a los agentes en el flujo de trabajo—desde Investigador, a agentes personalizados y agentes externos.
Sales Development Agent ayuda a escalar la capacidad de su equipo de ventas, a través de trabajar de manera autónoma en su nombre para ayudar a construir cartas de clientes, fomentar leads y personalizar el contacto, para que los vendedores puedan centrarse en cerrar operaciones. Se conecta con datos de ventas en herramientas CRM como Salesforce y Dynamics 365 y puede actuar directo en el flujo de trabajo en las aplicaciones de Microsoft 365. Disponible a través del Programa Frontier en diciembre de 2025.
El mes pasado, presentamos nuevas formas de crear aplicaciones, flujos de trabajo y agentes con facilidad, por medio de lenguaje natural. App Builder facilita la creación y despliegue de aplicaciones en minutos, y el agente Workflows ayuda a automatizar tareas según un calendario o en respuesta a eventos (disponible a través del programa Frontier).
Ponemos a disposición en Microsoft Copilot Studio una amplia gama de modelos OpenAI y Anthropic para crear agentes a través del modelo adecuado para la tarea. Y la semana pasada añadimos el último modelo de OpenAI, GPT-5.1 (la disponibilidad de modelos varía según el modelo).
Microsoft Agent Factory ayuda a las organizaciones a crear agentes por medio de Microsoft Foundry y Copilot Studio y a desplegarlos en cualquier lugar, incluido Microsoft 365 Copilot, con un único plan medido, sin necesidad de licencias ni provisiones iniciales. Las organizaciones elegibles también pueden acceder al apoyo práctico de los mejores Ingenieros Desplegados Avanzados de IA (FDEs, por sus siglas en inglés) y acceder a formación adaptada basada en roles para mejorar la fluidez en IA entre equipos. Disponible a nivel general.
El agente de Workforce Insights ofrece a líderes y gestores información completa y en tiempo real sobre su equipo y organización a través de roles, antigüedad y ubicación, para ayudar a los líderes a tomar decisiones basadas en datos sobre la fuerza laboral. Esto forma parte de un nuevo conjunto de agentes diseñados para mejorar la experiencia de los empleados al fomentar conexiones y acelerar el desarrollo de habilidades. Disponible a través del Programa Frontier.
Los agentes ya han comenzado a cambiar la manera en que las personas trabajan, e IDC predice que habrá 1.300 millones de agentes para 2028.1 Agent 365 es el plano de control para los agentes, que extiende la infraestructura en la que confían para gestionar a su personal a través de agentes.
Ya sea que sus agentes se creen por medio de Copilot Studio, Microsoft Foundry, frameworks de código abierto o a partir de un ecosistema creciente de socios, incluidos Adobe, Manus, SAP, ServiceNow, Workday y más, Agent 365 les ayuda a desplegarlos, organizarlos y gobernarlos de forma segura. Agent 365 incluye cinco capacidades clave:
Un registro que les proporcione una única fuente de verdad para todos los agentes de su organización.
Control de acceso para gestionar agentes y limitar su acceso solo a los recursos necesarios para realizar tareas específicas.
Visualización: un panel unificado y análisis avanzados para ver las conexiones entre agentes, personas y datos, y monitorizar el comportamiento y el rendimiento de los agentes en tiempo real.
Interoperabilidad para equipar a los agentes con aplicaciones y datos que simplifiquen los flujos de trabajo entre humanos y agentes.
Seguridad para ayudar a proteger a los agentes de amenazas y vulnerabilidades y detectar, investigar y remediar ataques que tienen como objetivo a los agentes.
Agent 365 une inteligencia, seguridad y gobernanza para liberar potentes capacidades agénticas con la seguridad en la que confías en Microsoft. Disponible a través del programa Frontier.
Preguntar a Copilot en la barra de tareas es su punto de entrada a Microsoft 365 Copilot, agentes y búsqueda—accesible al instante por voz o texto. Ahora pueden usar la barra de tareas de Windows para encontrar aplicaciones, archivos y configuraciones, y lanzar, invocar, monitorizar y gestionar agentes como Researcher, con aún más agentes propios y de terceros más adelante. En vista previa pública.
Windows 365 para Agentes permite a los creadores crear y desplegar agentes de nivel empresarial en ordenadores en la nube seguros y controlados por políticas, transmitidos desde Microsoft Cloud. Impulsa iniciativas avanzadas de IA de primera mano como Copilot Studio en un entorno Windows completo y Microsoft Researcher en un entorno Linux. Estamos entusiasmados por ver cómo los principales desarrolladores de agentes—incluidos Manus, Fellou, GenSpark, Simular y Tinyfish—ya desean aprovechar Windows 365 para Agentes y ofrecer soluciones de IA de próxima generación. Manus, por ejemplo, utiliza Windows 365 para ofrecer computación segura, escalable y siempre disponible para su agente general de IA, para permitir a los usuarios acceder a asistencia inteligente donde trabajen. Con Windows 365 para Agentes, las organizaciones pueden acelerar la adopción de la IA y escalar cargas de trabajo con las protecciones de cumplimiento que ustedes esperarían de Microsoft. En vista previa pública.
Otras noticias
Capacitamos a todos los profesionales de la seguridad con 12 nuevos agentes integrados en el flujo de trabajo en Microsoft Defender, Entra, Intune y Purview. Para facilitar más que nunca el aprovechamiento del poder de estos agentes, Microsoft Security Copilot se incluirá para todos los clientes de Microsoft 365 E5. Estos cambios comienzan a implementarse hoy para los clientes de Security Copilot con Microsoft 365 E5 y seguirán implementándose en todos los clientes con licencia Microsoft 365 E5 en los próximos meses. Más información aquí.
Estamos comprometidos a ayudar a cada organización a aprovechar el poder de la IA para crecer y escalar su negocio. Microsoft 365 Copilot Business es una nueva oferta para llevar a todas las IA de pequeñas y medianas empresas diseñadas para el trabajo. Microsoft 365 Copilot Business está disponible para empresas con menos de 300 usuarios por 21 dólares por usuario al mes, a partir de diciembre.*
Empiecen hoy mismo
Innovamos con Copilot y agentes para ayudar a cada cliente a convertirse en una Empresa Frontera. Visiten Microsoft365.com/copilot o descarguen la aplicación de Microsoft 365 en su dispositivo móvil. Para las últimas investigaciones y conocimientos sobre la IA en el trabajo, visiten WorkLab.
1IDC Info Snapshot, patrocinado por Microsoft, 1.300 millones de agentes de IA para mayo de 2025 #US53361825
Amazon CloudWatch Logs now supports automatically running Logs Insights queries on a recurring schedule for your log analysis needs. With scheduled queries, you can now automate log analysis tasks and deliver query results to Amazon S3 and Amazon EventBridge.
With today’s launch, you can track trends, monitor key operational metrics, and detect anomalies without needing to manually re-run queries or maintain custom automation. This feature makes it easier to maintain continuous visibility into your applications and infrastructure, streamline operational workflows, and ensure consistent insight generation at scale. For example, you can setup scheduled queries for your weekly audit reporting. The query results can also be stored in Amazon S3 for analysis, or trigger incident response workflows through Amazon EventBridge. The feature supports all CloudWatch Logs Insights query languages and helps teams improve operational efficiency by eliminating manual query executions.
Scheduled queries is available in US East (Ohio), US East (N. Virginia), US West (N. California), US West (Oregon), Asia Pacific (Mumbai), Asia Pacific (Osaka), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), Canada (Central), Europe (Frankfurt), Europe (Ireland), Europe (London), Europe (Paris), Europe (Stockholm), and South America (São Paulo).
You can configure a scheduled query using the Amazon CloudWatch console, AWS Command Line Interface (AWS CLI), AWS Cloud Development Kit (AWS CDK), and AWS SDKs. For more information, visit the Amazon CloudWatch documentation.
Amazon CloudWatch Logs now supports automatically running Logs Insights queries on a recurring schedule for your log analysis needs. With scheduled queries, you can now automate log analysis tasks and deliver query results to Amazon S3 and Amazon EventBridge.
With today’s launch, you can track trends, monitor key operational metrics, and detect anomalies without needing to manually re-run queries or maintain custom automation. This feature makes it easier to maintain continuous visibility into your applications and infrastructure, streamline operational workflows, and ensure consistent insight generation at scale. For example, you can setup scheduled queries for your weekly audit reporting. The query results can also be stored in Amazon S3 for analysis, or trigger incident response workflows through Amazon EventBridge. The feature supports all CloudWatch Logs Insights query languages and helps teams improve operational efficiency by eliminating manual query executions.
Scheduled queries is available in US East (Ohio), US East (N. Virginia), US West (N. California), US West (Oregon), Asia Pacific (Mumbai), Asia Pacific (Osaka), Asia Pacific (Seoul), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), Canada (Central), Europe (Frankfurt), Europe (Ireland), Europe (London), Europe (Paris), Europe (Stockholm), and South America (São Paulo).
You can configure a scheduled query using the Amazon CloudWatch console, AWS Command Line Interface (AWS CLI), AWS Cloud Development Kit (AWS CDK), and AWS SDKs. For more information, visit the Amazon CloudWatch documentation.
Amazon OpenSearch Service now includes Cluster Insights, a monitoring solution that provides comprehensive operational visibility of your clusters through a single dashboard. This eliminates the complexity of having to analyze and correlate various logs and metrics to identify potential risks to cluster availability or performance. The solution automates the consolidation of critical operational data across nodes, indices, and shards, transforming complex troubleshooting into a streamlined process.
When investigating performance issues like slow search queries, Cluster Insights displays relevant performance metrics, affected cluster resources, top-N query analysis, and specific remediation steps in one comprehensive view. The solution operates through OpenSearch UI’s resilient architecture, maintaining monitoring capabilities even during cluster unavailability. Users gain immediate access to account-level cluster summaries, enabling efficient management of multiple deployments.
Cluster Insights is available at no additional cost for OpenSearch version 2.17 or later in all Regions where OpenSearch UI is available. View the complete list of supported Regions here.
Amazon OpenSearch Service now includes Cluster Insights, a monitoring solution that provides comprehensive operational visibility of your clusters through a single dashboard. This eliminates the complexity of having to analyze and correlate various logs and metrics to identify potential risks to cluster availability or performance. The solution automates the consolidation of critical operational data across nodes, indices, and shards, transforming complex troubleshooting into a streamlined process. When investigating performance issues like slow search queries, Cluster Insights displays relevant performance metrics, affected cluster resources, top-N query analysis, and specific remediation steps in one comprehensive view. The solution operates through OpenSearch UI’s resilient architecture, maintaining monitoring capabilities even during cluster unavailability. Users gain immediate access to account-level cluster summaries, enabling efficient management of multiple deployments. Cluster Insights is available at no additional cost for OpenSearch version 2.17 or later in all Regions where OpenSearch UI is available. View the complete list of supported Regions here. To learn more about Cluster Insights, refer to our technical documentation.
Amazon ECR now offers a new archive storage class to reduce storage costs for large volumes of rarely accessed container images. The new archive storage class helps you meet your compliance and retention requirements while optimizing storage cost. As part of this launch, ECR lifecycle policies now support archiving images based on last pull time, allowing you to use lifecycle rules to automatically archive images based on usage patterns.
To get started, you can archive images by configuring lifecycle rules to automatically archive images based on criteria such as image age, count, or last pull time, or using the ECR Console or API to archive images individually. You can archive an unlimited number of images. Archived images do not count against your image per repository limit. Once the images are archived, they are no longer accessible for pulls, but can be easily restored via ECR Console, CLI, or API within 20 minutes. Once restored, images can be pulled normally. All archival and restore operations are logged through CloudTrail for auditability.
The new ECR archive storage class is available in all AWS Commercial and AWS GovCloud (US) Regions. For pricing, visit the pricing page. To learn more, visit the documentation.
Amazon ECR now offers a new archive storage class to reduce storage costs for large volumes of rarely accessed container images. The new archive storage class helps you meet your compliance and retention requirements while optimizing storage cost. As part of this launch, ECR lifecycle policies now support archiving images based on last pull time, allowing you to use lifecycle rules to automatically archive images based on usage patterns. To get started, you can archive images by configuring lifecycle rules to automatically archive images based on criteria such as image age, count, or last pull time, or using the ECR Console or API to archive images individually. You can archive an unlimited number of images. Archived images do not count against your image per repository limit. Once the images are archived, they are no longer accessible for pulls, but can be easily restored via ECR Console, CLI, or API within 20 minutes. Once restored, images can be pulled normally. All archival and restore operations are logged through CloudTrail for auditability. The new ECR archive storage class is available in all AWS Commercial and AWS GovCloud (US) Regions. For pricing, visit the pricing page. To learn more, visit the documentation.
Today, AWS announces the general availability of Amazon Elastic Compute Cloud (Amazon EC2) P6-B300 instances, accelerated by NVIDIA Blackwell Ultra B300 GPUs. Amazon EC2 P6-B300 instances provide 8x NVIDIA Blackwell Ultra GPUs with 2.1 TB high bandwidth GPU memory, 6.4 Tbps EFA networking, 300 Gbps dedicated ENA throughput, and 4 TB of system memory.
P6-B300 instances deliver 2x networking bandwidth, 1.5x GPU memory size, and 1.5x GPU TFLOPS (at FP4, without sparsity) compared to P6-B200 instances, making them well suited to train and deploy large trillion-parameter foundation models (FMs) and large language models (LLMs) with sophisticated techniques. The higher networking and larger memory deliver faster training times and more token throughput for AI workloads.
P6-B300 instances are now available in the p6-b300.48xlarge size through Amazon EC2 Capacity Blocks for ML and Savings Plans in the following AWS Region: US West (Oregon). For on-demand reservation of P6-B300 instances, please reach out to your account manager.
Today, AWS announces the general availability of Amazon Elastic Compute Cloud (Amazon EC2) P6-B300 instances, accelerated by NVIDIA Blackwell Ultra B300 GPUs. Amazon EC2 P6-B300 instances provide 8x NVIDIA Blackwell Ultra GPUs with 2.1 TB high bandwidth GPU memory, 6.4 Tbps EFA networking, 300 Gbps dedicated ENA throughput, and 4 TB of system memory.
P6-B300 instances deliver 2x networking bandwidth, 1.5x GPU memory size, and 1.5x GPU TFLOPS (at FP4, without sparsity) compared to P6-B200 instances, making them well suited to train and deploy large trillion-parameter foundation models (FMs) and large language models (LLMs) with sophisticated techniques. The higher networking and larger memory deliver faster training times and more token throughput for AI workloads.
P6-B300 instances are now available in the p6-b300.48xlarge size through Amazon EC2 Capacity Blocks for ML and Savings Plans in the following AWS Region: US West (Oregon). For on-demand reservation of P6-B300 instances, please reach out to your account manager.
To learn more about P6-B300 instances, visit Amazon EC2 P6 instances.
Today, Amazon Bedrock introduces two new inference service tiers to optimize costs and performance for different AI workloads. The new Flex tier offers cost-effective pricing for non-time-critical applications like model evaluations and content summarization while the Priority tier provides premium performance and preferential processing for mission-critical applications. For most models that support Priority Tier, customers can realize up to 25% better output tokens per second (OTPS) latency compared to standard tier. These join the existing Standard tier for everyday AI applications with reliable performance.
These service tiers address key challenges that organizations face when deploying AI at scale. The Flex tier is designed for non-interactive workloads that can tolerate longer latencies, making it ideal for model evaluations, content summarization, labeling and annotation, and multistep agentic workflow, and it’s priced at a discount relative to the Standard tier. During periods of high demand, Flex requests receive lower priority relative to the Standard tier. The Priority tier is an ideal fit for mission critical applications, real-time end-user interactions, and interactive experiences where consistent, fast responses are essential. During periods of high demand, Priority requests receive processing priority, at a premium price, over other service tiers. These new service tiers are available today for a range of leading foundation models, including OpenAI (gpt-oss-20b, gpt-oss-120b), DeepSeek (DeepSeek V3.1), Qwen3 (Coder-480B-A35B-Instruct, Coder-30B-A3B-Instruct, 32B dense, Qwen3-235B-A22B-2507), and Amazon Nova (Nova Pro and Nova Premier). With these new options, Amazon Bedrock helps customers gain greater control over balancing cost efficiency with performance requirements, enabling them to scale AI workloads economically while ensuring optimal user experiences for their most critical applications.
For more information about the AWS Regions where Amazon Bedrock Priority and Flex inference service tiers are available, see the AWS Regions table
Today, Amazon Bedrock introduces two new inference service tiers to optimize costs and performance for different AI workloads. The new Flex tier offers cost-effective pricing for non-time-critical applications like model evaluations and content summarization while the Priority tier provides premium performance and preferential processing for mission-critical applications. For most models that support Priority Tier, customers can realize up to 25% better output tokens per second (OTPS) latency compared to standard tier. These join the existing Standard tier for everyday AI applications with reliable performance.
These service tiers address key challenges that organizations face when deploying AI at scale. The Flex tier is designed for non-interactive workloads that can tolerate longer latencies, making it ideal for model evaluations, content summarization, labeling and annotation, and multistep agentic workflow, and it’s priced at a discount relative to the Standard tier. During periods of high demand, Flex requests receive lower priority relative to the Standard tier. The Priority tier is an ideal fit for mission critical applications, real-time end-user interactions, and interactive experiences where consistent, fast responses are essential. During periods of high demand, Priority requests receive processing priority, at a premium price, over other service tiers. These new service tiers are available today for a range of leading foundation models, including OpenAI (gpt-oss-20b, gpt-oss-120b), DeepSeek (DeepSeek V3.1), Qwen3 (Coder-480B-A35B-Instruct, Coder-30B-A3B-Instruct, 32B dense, Qwen3-235B-A22B-2507), and Amazon Nova (Nova Pro and Nova Premier). With these new options, Amazon Bedrock helps customers gain greater control over balancing cost efficiency with performance requirements, enabling them to scale AI workloads economically while ensuring optimal user experiences for their most critical applications.
For more information about the AWS Regions where Amazon Bedrock Priority and Flex inference service tiers are available, see the AWS Regions table
Learn more about service tiers in our News Blog and documentation.
AWS Lambda now supports creating serverless applications using Python 3.14. Developers can use Python 3.14 as both a managed runtime and a container base image, and AWS will automatically apply updates to the managed runtime and base image as they become available.
Python 3.14 is the latest long-term support release of Python. This release provides Lambda customers access to the latest Python 3.14 language features. You can use Python 3.14 with Lambda@Edge (in supported Regions), allowing you to customize low-latency content delivered through Amazon CloudFront. Powertools for AWS Lambda (Python), a developer toolkit to implement serverless best practices and increase developer velocity, also supports Python 3.14. You can use the full range of AWS deployment tools, including the Lambda console, AWS CLI, AWS Serverless Application Model (AWS SAM), AWS CDK, and AWS CloudFormation to deploy and manage serverless applications written in Python 3.14.
The Python 3.14 runtime is available in all Regions, including the AWS GovCloud (US) Regions and China Regions.
For more information, including guidance on upgrading existing Lambda functions, read our blog post. For more information about AWS Lambda, visit the product page.
AWS Lambda now supports creating serverless applications using Python 3.14. Developers can use Python 3.14 as both a managed runtime and a container base image, and AWS will automatically apply updates to the managed runtime and base image as they become available. Python 3.14 is the latest long-term support release of Python. This release provides Lambda customers access to the latest Python 3.14 language features. You can use Python 3.14 with Lambda@Edge (in supported Regions), allowing you to customize low-latency content delivered through Amazon CloudFront. Powertools for AWS Lambda (Python), a developer toolkit to implement serverless best practices and increase developer velocity, also supports Python 3.14. You can use the full range of AWS deployment tools, including the Lambda console, AWS CLI, AWS Serverless Application Model (AWS SAM), AWS CDK, and AWS CloudFormation to deploy and manage serverless applications written in Python 3.14. The Python 3.14 runtime is available in all Regions, including the AWS GovCloud (US) Regions and China Regions. For more information, including guidance on upgrading existing Lambda functions, read our blog post. For more information about AWS Lambda, visit the product page.
Amazon OpenSearch Serverless now supports detailed audit logging of data plane requests via AWS CloudTrail. This feature enables customers to record user actions on their collections, helping meet compliance regulations, improve security posture, and provide evidence for security investigations. Customers can now track user activities such as authorization attempts, index modifications, and search queries.
Customers can use CloudTrail to configure filters for OpenSearch Serverless collections with read-only and write-only options, or use advanced event selectors for more granular control over logged data events. All OpenSearch Serverless data events are delivered to an Amazon S3 bucket and optionally to Amazon CloudWatch Events, creating a comprehensive audit trail. This enhanced visibility into when and who made API calls helps security and operations teams monitor data access and respond to events in real-time.
Once configured with CloudTrail, audit logs will be continuously streamed with no additional customer action required. Audit Logs will be continuously streamed to CloudTrail and can be further analyzed there.
Amazon OpenSearch Serverless now supports detailed audit logging of data plane requests via AWS CloudTrail. This feature enables customers to record user actions on their collections, helping meet compliance regulations, improve security posture, and provide evidence for security investigations. Customers can now track user activities such as authorization attempts, index modifications, and search queries. Customers can use CloudTrail to configure filters for OpenSearch Serverless collections with read-only and write-only options, or use advanced event selectors for more granular control over logged data events. All OpenSearch Serverless data events are delivered to an Amazon S3 bucket and optionally to Amazon CloudWatch Events, creating a comprehensive audit trail. This enhanced visibility into when and who made API calls helps security and operations teams monitor data access and respond to events in real-time. Once configured with CloudTrail, audit logs will be continuously streamed with no additional customer action required. Audit Logs will be continuously streamed to CloudTrail and can be further analyzed there. Please refer to the AWS Regional Services List for more information about Amazon OpenSearch Service availability. To learn more about OpenSearch Serverless, see the documentation.
Today, EC2 Auto Scaling is launching a new API, LaunchInstances, which gives customers more control and flexibility over how EC2 Auto Scaling provisions instances while providing instant feedback on capacity availability.
Customers use EC2 Auto Scaling for automated fleet management. With scaling policies, EC2 Auto Scaling can automatically add instances when demand spikes and remove them when traffic drops, ensuring customers’ applications always have the right amount of compute. EC2 Auto Scaling also offers the ability to monitor and replace unhealthy instances. In certain use cases, customers may want to specify exactly where EC2 Auto Scaling should launch additional instances and need immediate feedback on capacity availability. The new LaunchInstances API allows customers to precisely control where instances are launched by specifying an override for any Availability Zone and/or subnet in an Auto Scaling group, while providing immediate feedback on capacity availability. This synchronous operation gives customers real-time insight into scaling operations, enabling them to quickly implement alternative strategies if needed. For additional flexibility, the API includes optional asynchronous retries to help reach the desired capacity.
This feature is now available in US East (N. Virginia), US West (Oregon), Europe (Ireland), and Asia Pacific (Singapore), at no additional cost beyond standard EC2 and EBS usage. To get started, visit the AWS Command Line Interface (CLI) and the AWS SDKs. To learn more about this feature, visit the AWS documentation.
Today, EC2 Auto Scaling is launching a new API, LaunchInstances, which gives customers more control and flexibility over how EC2 Auto Scaling provisions instances while providing instant feedback on capacity availability. Customers use EC2 Auto Scaling for automated fleet management. With scaling policies, EC2 Auto Scaling can automatically add instances when demand spikes and remove them when traffic drops, ensuring customers’ applications always have the right amount of compute. EC2 Auto Scaling also offers the ability to monitor and replace unhealthy instances. In certain use cases, customers may want to specify exactly where EC2 Auto Scaling should launch additional instances and need immediate feedback on capacity availability. The new LaunchInstances API allows customers to precisely control where instances are launched by specifying an override for any Availability Zone and/or subnet in an Auto Scaling group, while providing immediate feedback on capacity availability. This synchronous operation gives customers real-time insight into scaling operations, enabling them to quickly implement alternative strategies if needed. For additional flexibility, the API includes optional asynchronous retries to help reach the desired capacity. This feature is now available in US East (N. Virginia), US West (Oregon), Europe (Ireland), and Asia Pacific (Singapore), at no additional cost beyond standard EC2 and EBS usage. To get started, visit the AWS Command Line Interface (CLI) and the AWS SDKs. To learn more about this feature, visit the AWS documentation.
