AWS AgentCore Runtime V2 reclaims cold memory and uses snapshots for predictable agent starts
The managed runtime changes memory accounting and startup so long-running and interactive agents can scale with less idle footprint.
Key takeaways
AWS announced the new AgentCore Runtime on September 18. The service now starts sessions with a smaller resident memory profile, pages memory in when touched, and reclaims released or cold memory instead of charging against the session's high-water mark. When a runtime is created or updated, AgentCore initializes the container once, captures a compact snapshot, and restores later instances from that state. AWS reports 5,000 cold invocations per agent across five image sizes under default quotas, with roughly two-second P75 starts from 200 MB through 2 GB images. Teams opt in with platformVersion V2 and should retest latency, memory, and cost using their own workloads.
Direct answer
AWS announced the new AgentCore Runtime on September 18. The service now starts sessions with a smaller resident memory profile, pages memory in when touched, and reclaims released or cold memory instead of charging against the session's high-water mark. When a runtime is created or updated, AgentCore initializes the container once, captures a compact snapshot, and restores later instances from that state. AWS reports 5,000 cold invocations per agent across five image sizes under default quotas, with roughly two-second P75 starts from 200 MB through 2 GB images. Teams opt in with platformVersion V2 and should retest latency, memory, and cost using their own workloads.
Verified facts
The new runtime starts with a smaller resident footprint, pages memory in on demand, and reclaims released or cold session memory.
AgentCore completes initialization and health checks when a runtime is created or updated, then captures a compact snapshot for future instance restores.
AWS reports 5,000 cold invocations per agent across five image sizes, with about two-second P75 cold starts from 200 MB through 2 GB images.
What changed
- Session usage no longer stays pinned to peak memory
- Instances restore from a compact snapshot after one initialization
- Container image size has much less influence on cold-start time
- V2 is selected explicitly through the platform version
Impact for AI users
More predictable starts can reduce the first-turn delay for interactive agents, while cold-memory reclamation may lower idle footprint for long-running or event-driven work. AWS's figures come from its test conditions and do not replace production capacity planning. Before migration, teams should hold the model, image, concurrency, and request distribution constant and compare P75, P95, memory GB-hours, and cost per completed task.
Operating checklist
- Clone the production image and establish a V1 baseline
- Enable V2 under the same concurrency and request distribution
- Record startup, memory, failure rate, and completed-task cost
- Expand traffic only after every stop condition passes
AI frontier news and analysis, AI software and model tools, AI skill tutorials and validation methods, and AI account and permission guidance
FAQ
Does every image have a guaranteed two-second start?
No. About two seconds is the reported P75 in AWS's test, and real workloads require independent measurement.
How is the new runtime enabled?
AWS instructs users to set platformVersion to V2 when creating or updating a runtime.
Summary
Runtime V2 changes memory and startup mechanics in measurable ways. Production adoption should be decided by like-for-like workload evidence, not announcement numbers.
This AI-assisted article is checked by ENHE AI automation for official sources, bilingual fields, media rights, page safety, and historical duplication before publication.
What this means for everyday users
More predictable starts can reduce the first-turn delay for interactive agents, while cold-memory reclamation may lower idle footprint for long-running or event-driven work. AWS's figures come from its test conditions and do not replace production capacity planning. Before migration, teams should hold the model, image, concurrency, and request distribution constant and compare P75, P95, memory GB-hours, and cost per completed task.
Tools you may use
Related tutorials
Related Tools And Tutorials
Use the following ENHE AI sections to continue from the news signal into tool selection, account-service guidance, or practical learning.
Related reading
SageMaker Studio Simplifies New-Customer Access: Credits, Setup, and Cost Boundaries
AWS announced a simplified path into Amazon SageMaker Studio for new customers on October 19, 2026. Eligible users can sign up with Amazon, Apple, GitHub, Google, or email in two or three steps, and most do not need to enter a credit card first. Studio is created in about 20 seconds. New customers receive $100 in Free Tier credits and may earn another $100; the free period ends after six months or when credits are exhausted, and resource usage then stops automatically. This article provides a bounded first-experiment checklist covering account eligibility, GPU quota visibility, endpoint shutdown, credit monitoring, and the decision point before a paid upgrade.
Smartsheet MCP Is Generally Available for Amazon Quick and Kiro Workflows
AWS has published a guide to the generally available Smartsheet MCP Server, which exposes more than forty tools for reading, writing, and managing sheets, rows, columns, and automations. Business users can connect through Amazon Quick, while developers can configure the server in Kiro. This article compares the two paths, lists the documented account, authentication, and regional requirements, and proposes a safe first task: identify blocked work and produce a review checklist without writing changes. Only after results, user permissions, and organization controls are verified should a team enable a narrowly scoped write action. Natural-language requests do not grant extra access, and a connected agent should not be treated as an unrestricted project administrator.
When Should an AI Agent Act Autonomously? A Scope, Know, Bound, Observe Checklist
An AI agent should receive more autonomy only when its actions can be verified, not because one demonstration looked successful. AWS describes a Trust in Depth framework organized around four lifecycle questions: Scope, Know, Bound, and Observe. This durable guide turns those questions into an operational checklist for a small team. Using an internal task-status update as the example, it identifies evidence required before launch, controls during execution, records after execution, and conditions that should stop the agent and request human review. Precise rules can sometimes be tested or formally checked, while tone, sensitive context, and high-impact decisions still require judgment. The objective is bounded, reviewable autonomy rather than either constant approval or unrestricted action.
Reflection Beam Arrives on AWS: How to Choose an Agentic Coding Deployment
AWS introduced Reflection AI's Beam on October 16, 2026 as an open-weight model for agentic coding and long-horizon work. Beam is designed to resolve software issues, modify code, call tools, interpret intermediate results, and continue multi-step workflows. AWS describes a sparse mixture-of-experts architecture with 501 billion total parameters, 23 billion active parameters, and an effective one-million-token context length reported by Reflection. Weights are released under Apache 2.0 in BF16, FP8, and NVFP4 variants. This guide separates the managed SageMaker JumpStart path from the more controllable HyperPod route and explains the operational evidence teams should collect before trusting either in real projects.
Nine-Stage AI Agent Evaluation Checklist: From First Commit to Production Drift
AWS published a nine-stage production AI-agent evaluation framework built around Amazon Bedrock AgentCore on October 16, 2026. It follows an agent from defining quality and comparing experiments through merge checks, deployment gates, canary scoring, adversarial testing, drift detection, live pulse monitoring, and feedback-driven improvement. The framework evaluates five dimensions—safety, quality, performance, cost, and operational control—at tool, individual-agent, and multi-agent levels. This durable guide converts the public framework into a vendor-neutral checklist for ordinary teams, with minimum evidence, ownership, and stop conditions designed to catch soft failures where latency and uptime remain normal while answer quality quietly declines over time.
A Banking Agent Writes Then Reverses: AWS Describes a Demo, Not Production Trading
An October 5 AWS post describes Texas Capital Bank’s Trust-Earned Autonomy pattern, in which reviewer agreement can inform an agent’s permissions and a prior write can be reversed if trust declines. AWS explicitly says the agent pattern is still a demonstration running on demo-scale Apache Fineract with synthetic data, not a production workload or real production transaction flow. The bank’s separate document-extraction capability is described as operating in production. This article focuses on revocable authority, auditability, and system boundaries, and avoids presenting the demonstration as an already deployed autonomous banking agent. Teams should check applicability against their own environment, record assumptions, and validate decisions with representative operational evidence.