AWS Launches Bedrock AgentCore: A Scalable Platform for Enterprise AI Agents with Open Source Support

On July 16, 2025, cloud computing giant Amazon Web Services (AWS) unveiled Amazon Bedrock AgentCore, a new enterprise-grade platform designed to build, deploy, and operate AI agents securely and at scale. The announcement came during AWS's New York Summit, where Swami Sivasubramanian, AWS Vice President of Agentic AI, emphasized the platform’s ability to help organizations transition from experimental to production-ready AI agent systems, trusted for critical business processes. Deepak Singh, AWS Vice President of Databases and AI, highlighted the blurring lines between traditional applications and AI agents, suggesting that AgentCore will be instrumental in this transformation. The platform is a modular stack, offering essential services to move AI agents from prototypes to fully functional tools. These services include: AgentCore Runtime: A serverless, low-latency execution environment supporting multimodal workloads and ensuring session isolation. AgentCore Memory: Both long- and short-term memory services that enable agents to learn from past interactions and maintain contextual knowledge. AgentCore Identity: OAuth-based identity and access management, facilitating secure interactions with systems like GitHub, Slack, and Salesforce. AgentCore Observability: Built-in dashboards, debugging, and telemetry tools compatible with OpenTelemetry, LangSmith, and Datadog. AgentCore Gateway: Converts internal APIs, Lambda functions, and third-party services into agent-compatible tools using the Model Context Protocol (MCP). AgentCore Browser: Provides headless browser access for agents to interact with websites autonomously. AgentCore Code Interpreter: A secure environment for executing code generated by agents, aiding in analysis and visualization. AWS’s approach with AgentCore is notable for its flexibility and interoperability. The platform supports any agent framework or foundational model, not just those hosted within AWS. This includes compatibility with open-source toolkits like CrewAI, LangChain, LlamaIndex, LangGraph, and AWS’s own Strands Agents SDK. Several early adopters have already begun integrating AgentCore into their operations. Cloud storage company Box is using the platform to enhance its content management tools with top-tier security and compliance, aiming to scale AI capabilities for its enterprise customers. Ben Kus, Box’s CTO, lauded the integration’s potential to deliver robust and secure AI-driven solutions. Brazil’s largest bank, Itaú Unibanco, is leveraging AgentCore to develop hyper-personalized digital banking experiences. Carlos Eduardo Mazzei, the bank’s CTO, praised the platform’s role in delivering intuitive and efficient banking services. Innovaccer, a healthcare technology provider, has built the Healthcare Model Context Protocol (HMCP) on top of AgentCore Gateway, converting existing APIs into agent-compatible tools. CEO and co-founder Abhinav Shashank described Gateway as a “game-changer” for healthcare, enhancing scalability, trust, and operational efficiency. Meanwhile, marketing firm Epsilon is using AgentCore to accelerate campaign build times and improve customer engagement, expecting reductions of up to 30% in build times and enhanced personalization. Availability and Pricing: AgentCore is currently available in preview in select AWS regions, including US East (N. Virginia), US West (Oregon), Asia Pacific (Sydney), and Europe (Frankfurt). Developers can use it for free until September 16, 2025. After this period, pricing will switch to a consumption-based model with no upfront commitments or minimum fees. Key pricing details: - Runtime, Browser, and Code Interpreter: Priced per second, with rates at $0.0895 per vCPU-hour and $0.00945 per GB-hour. - Gateway: Charges $0.005 per 1,000 tool API invocations, $0.025 per 1,000 search queries, and $0.02 per 100 tools indexed per month. - Memory: Costs $0.25 per 1,000 short-term memory events and $0.75 per 1,000 long-term memories stored (or $0.25 with custom strategies), along with $0.50 per 1,000 retrievals. - Identity: Included at no extra charge when used via Runtime or Gateway, otherwise $0.010 per 1,000 token or API key requests. - Observability: Billed via Amazon CloudWatch rates. Industry Impact: The launch of AgentCore marks a significant step in the emerging field of enterprise AI agents. It positions AWS as a leader in a highly competitive market, alongside offerings from OpenAI, Google, and startups like Cognition. Industry insiders view the platform as a critical tool for democratizing AI agent development, making it easier for enterprises to harness the power of AI without the need for extensive proprietary resources. By supporting open-source frameworks and ensuring flexible pricing, AWS aims to foster innovation and adoption across various sectors. Company Profile: Amazon Web Services (AWS) is a subsidiary of Amazon, providing a comprehensive suite of cloud computing services. With AgentCore, AWS continues its tradition of delivering scalable and secure solutions, addressing the growing demand for AI-driven tools in the enterprise market. This latest platform underscores AWS's commitment to staying at the forefront of technological advancements and helping businesses navigate the complexities of AI integration.