█

AI/TLDR

Buy me a coffee
Buy me a coffee

AWS Emulator

A free local stand-in for the AWS APIs, for tests, CI and AI coding agents

Code Sandboxes & IsolationOpen source
Updated
11 Nov 2026
Language
Rust
License
CC0-1.0
Coverage
1 story

What's new

11 Nov 2026

AWS launched AWS Emulator, a free local emulator of the AWS APIs for Linux, Docker and Rust tests, and published its CC0 source and fidelity tests on GitHub.

Latest news

IN SHORT

What is AWS Emulator?

AWS Emulator is a local emulator of the AWS APIs, published by Amazon Web Services. It sits in the Code Sandboxes & Isolation category of the Agents & Automation tools and is open-source, written in Rust and licensed CC0-1.0. To get started: install the package (Debian, amd64).

Updated Markdown version ↗

Overview

AWS Emulator is a local emulator of the AWS APIs, published by Amazon Web Services. You run it on your own machine, a development server or a CI runner, set a few environment variables, and the AWS CLI and AWS SDKs send their requests to emulated resources instead of to AWS. Your code does not change, you do not need an AWS account, and there is no service fee. It has no graphical interface: it is a single endpoint that speaks the AWS APIs.

The emulator is written in Rust and serves HTTPS on 127.0.0.1:9297 by default. On each start it generates fresh access keys, a config file and a certificate under ~/.aws/emulator/<port>/, and it checks SigV4 signatures like AWS does, so only callers that use those keys get through. All resources live in memory and are deleted when the emulator stops, which gives every test run a clean state. Calls to operations it has not implemented return HTTP 501, so tests fail loudly instead of passing by accident.

Diagram: test writers, issue finders and fixer agents working through one shared CodeCommit repository
Three teams of agents build AWS Emulator: test writers, issue finders and fixersAbout Amazon ↗

The project is also an experiment in agent-built software. Amazon says three teams of AI agents build and maintain it from the AWS API models, with real AWS as the reference: test writers verify scenarios against AWS, issue finders replay those fidelity tests against the emulator and file failures, and fixers merge one fix per issue through an automated gate for build, review, security and fidelity. AWS says its engineers have not written or reviewed the code. Each release publishes per-service fidelity scores, and the tests themselves are public.

What it does

  • Works with the AWS CLI and AWS SDKs through environment variables, with no code changes
  • One local HTTPS endpoint for every emulated service, with SigV4 checks and per-run keys
  • In-memory state that is wiped on stop, plus several emulators side by side on different ports
  • Native RPM and Debian packages for x86_64 and aarch64 Linux; Docker on macOS and Windows
  • Embedded mode for Rust tests (aws-emulator-embedded), with separate state per instance
  • An aws-emulator skill that lets AI coding agents install, start and connect the emulator
  • Per-service fidelity scores with each release and a public fidelity test suite
  • No service fee; CC0-1.0 source on GitHub (Apache-2.0 for vendor/)
Diagram: fixes batch from mainline through build, test and final checks to a release branch, GitHub and a binary
The release gate re-verifies batched fixes before a release is packaged and published to GitHubAbout Amazon ↗

Getting started

On Linux, install the package, start the emulator in one terminal and point your shell at it from another. On macOS and Windows, follow the Docker guide in the user guide instead.

Install the package (Debian, amd64)

RPM packages for dnf and arm64 builds are also published; the user guide lists each command.

bashbash
curl -fsSLo aws-emulator.deb https://downloads.aws.com/aws-emulator/packages/aws-emulator/LTS/latest/aws-emulator_amd64.deb
sudo apt-get install -y ./aws-emulator.deb
aws-emulator -v

Start the emulator

It keeps running in this terminal and serves HTTPS on 127.0.0.1:9297. Use -p to pick another port.

bashbash
aws-emulator

Point the AWS CLI and SDKs at it

In a second terminal, set the endpoint, the emulator's credential, config and certificate files, and unset any real credentials.

bashbash
export AWS_ENDPOINT_URL=https://127.0.0.1:9297
export AWS_SHARED_CREDENTIALS_FILE=~/.aws/emulator/9297/emulator_credentials
export AWS_CONFIG_FILE=~/.aws/emulator/9297/config
export AWS_CA_BUNDLE=~/.aws/emulator/9297/ca.pem
export AWS_EC2_METADATA_DISABLED=true
export AWS_DISABLE_HOST_PREFIX_INJECTION=true
export AWS_REGION=us-east-1
export AWS_DEFAULT_REGION=us-east-1
unset AWS_ACCESS_KEY_ID AWS_SECRET_ACCESS_KEY AWS_SESSION_TOKEN AWS_PROFILE

Check you are talking to the emulator

The emulator uses the single account 000000000000. If you see your own account ID, your requests are going to AWS.

bashbash
aws sts get-caller-identity
aws s3api create-bucket --bucket amzn-s3-demo-bucket
aws s3api list-buckets --query 'Buckets[].Name'

Or let a coding agent set it up

With AWS CLI 2.35.0 or later, add the aws-emulator skill so an AI coding agent can install, start and connect the emulator itself.

bashbash
aws agent-toolkit add-skill --skill-name aws-emulator --region us-east-1

Commands and code are distilled from the project's own documentation — always check the official repo for the latest.

When to use it

  • Run integration tests against S3, DynamoDB, SQS and other services on a laptop without an AWS account
  • Give CI pipelines AWS-shaped tests with no credentials and no cloud resources to clean up
  • Let an AI coding agent build and try AWS code without handing it real credentials
  • Check validation errors and API behaviour before deploying, then restart for a clean state

How AWS Emulator compares

AWS Emulator alongside other open-source code sandboxes & isolation tools AI/TLDR tracks, ranked by GitHub stars.

ToolStarsWhat it does
Daytona★ 71.3kDaytona is an open-source runtime that spins up isolated sandboxes in under 90ms so agents can safely run and persist AI-generated code.
NVIDIA NemoClaw★ 23kNVIDIA's reference stack for running OpenClaw, Hermes and LangChain Deep Agents Code inside OpenShell sandboxes, adding managed inference, network policy, snapshots and CLI lifecycle control.
OpenShell★ 21.4kNVIDIA's open-source runtime that runs autonomous AI agents in sandboxes, enforcing a declared policy on every file access, syscall and network connection and formally verifying policy changes.
OpenSandbox★ 17.2kOpenSandbox gives AI agents a safe place to run code and commands, with one unified API across Docker and Kubernetes runtimes and SDKs in five languages.
E2B★ 14.8kE2B is open-source infrastructure that runs AI-generated code inside secure, isolated cloud sandboxes, controlled from JavaScript or Python SDKs.
AX (Agent Executor)★ 14.6kGoogle's declarative orchestrator for agent workloads: a Task runs untrusted agent code in a sandbox, while Workspace, Gateway and Model resources wire up its repos, network allowlist and LLM credentials.
Astrid★ 10.3kA portable Rust runtime that executes software as sandboxed WebAssembly capsules, where every file, network, process and tool call is gated by a signed, revocable, per-principal capability instead of ambient authority.
AWS Emulator—A free local stand-in for the AWS APIs, for tests, CI and AI coding agents