Developers can use Amazon Bedrock AgentCore services to build foundational infrastructure for deploying and operating AI agents at scale.
AWS has launched a preview of Amazon Bedrock AgentCore, a set of enterprise-grade services developers can use to securely deploy and operate AI agents at scale using any framework and model.
Announced July 16, AgentCore services can be hosted on AWS’s Amazon Bedrock managed service or elsewhere. With this initiative, AWS says it is seeking to spare AI agent developers and engineers months spent building foundational infrastructure for identity controls, session management, memory systems, and observability.
Amazon Bedrock AgentCore lets users deploy highly capable AI agents securely, according to AWS. It offers infrastructure purpose-built for dynamic agent workloads, tools to enhance agents, and controls for real-world deployment. AgentCore can be used for applications such as customer support, workflow automation, and innovative AI-powered experiences. Services include:
- AgentCore Runtime: Provides low-latency serverless environments with session isolation, supporting any agent framework, including popular open source frameworks, tools, and models. Also handles multimodal workloads and long-running agents.
- AgentCore Memory: For managing session and long-term memory, providing relevant context to models while helping agents learn from past interactions.
- AgentCore Observability: For step-by-step visualization of agent execution with metadata tagging, custom scoring, trajectory inspection, and troubleshooting and debugging filters.
- AgentCore Identity: Enables AI agents to securely access AWS services and third-party tools and services such as GitHub, Salesforce, and Slack, either on behalf of users or by themselves, with pre-authorized user consent.
- AgentCore Gateway: Transforms existing APIs and AWS Lambda functions into agent-ready tools, providing unified access across protocols, including MCP (Model Context Protocol), and runtime discovery.
- AgentCore Browser: Providing managed browser instances to scale agents’ web automation workflows.
- AgentCore Code Interpreter: An isolated environment to run code that agents generate.
AgentCore services can be used together or independently and work with frameworks including CrewAI, LangGraph, LlamaIndex, and Strands Agents, as well as any foundation model inside or outside of Amazon Bedrock. Developers can try AgentCore services at no charge until September 16, 2025. Standard AWS pricing applies to any additional AWS services when using AgentCore. Users will be billed for AgentCore service usage beginning September 17, 2025.
