Amazon Web Services has launched runtime instances for Bedrock AgentCore, a new compute option that allows AI agents to run for up to fourteen days on managed EC2 infrastructure within customer accounts. The feature breaks through the eight-hour ceiling that constrained the platform's original serverless microVM sessions, targeting development teams whose agents require longer execution windows, GPU acceleration, or tighter collaboration among multiple agents sharing a host. According to principal developer advocate Sebastien Stormacq, runtime instances deliver persistent EC2-backed sessions with shared file systems, GPU-accelerated instance types, and support for Python and container images, while preserving the existing AgentCore APIs, identity model, and observability tools.

Runtime instances run on standard Amazon EC2 in the customer's account and are billed at typical EC2 rates plus a management fee, with the compute portion eligible for existing EC2 commitment discounts like Savings Plans and Reserved Instances. A cost analysis from eCorpIT suggests the break-even point sits at roughly 24% sustained CPU utilization before applying any Savings Plan discounts. The serverless microVM runtime continues to charge per vCPU hour and per GB hour with no separate management fee, and can prove cheaper for agents with low sustained CPU utilization and long idle periods. Runtime instances allow multiple agents to be deployed into a single runtime and collaborate on the same host through a shared session directory, rather than calling each other's APIs for every handoff. Teams can bring frameworks such as CrewAI, LangGraph, LlamaIndex, and Strands without changing their packaging model, using a simple decorator and either a zip file or container image.

Stormacq notes that "runtime instances handles all of that for you while integrating with the same AgentCore APIs, identity controls, and observability you already use with AgentCore Runtime microVMs." He also highlights that "your agents can call each other as tools within a shared session, iterating autonomously until the job is done." An engineering guide from Enkompass stresses that runtime instances are best suited to workloads that need longer than eight hours of continuous runtime, require GPUs or large memory footprints, or benefit from several agents co-located on the same host for tight collaboration. The microVM-based runtime remains the default choice for short, bursty, request-response style agents because it starts quickly, isolates sessions, and bills per second on actual CPU and peak memory up to eight hours.

AWS positions runtime instances as complementary to microVMs rather than a replacement, and encourages mixed topologies where an orchestrator agent on microVMs dispatches long-running work to workers on instances. The Enkompass guide expects most organizations to run both models side by side, placing high-volume, short-lived traffic on microVMs and reserving instances for a smaller set of long-running, stateful, or accelerator-heavy workloads that previously forced teams to maintain separate EC2 fleets. A capacity provider primitive underpins runtime instances, defining allowed instance families, operating system, networking, and storage, and acting as a contract between agents and the EC2 capacity that AgentCore will provision, patch, and scale on their behalf. Teams set bounds such as minimum and maximum instance counts and target utilization, then attach one or more agent runtimes to that provider with a session time-to-live of up to fourteen days, avoiding the need to manage Auto Scaling groups, launch templates, or AMI pipelines directly. The article argues that co-locating several agents on a single host and stopping sessions when work completes are the main levers for keeping instance-based costs under control. For teams already managing complex multi-agent workflows, the trade-off between upfront instance costs and operational overhead may tip decisively toward managed runtimes as orchestration complexity scales. Enterprises accustomed to fine-tuning infrastructure spend will need to weigh convenience against the incremental management fee that sits outside traditional discount programs.