Developer Tools·Standings
Chaos Engineering Software Compared
Chaos engineering platforms inject controlled failures into applications, hosts, containers, and cloud workloads to reveal resilience problems. Compare where a tool runs, what it connects to, its API or open-source status, free access, and cost.
The signal score out of 10 reflects usage, public API, documentation, open-source code, and a free tier. Check the product details for Kubernetes support, cloud and network faults, scheduling, deployment model, and blast-radius controls.
- 25 platforms
- 2 with a public API on this page
- 18 with a free tier
- Checked 23 Sept 2026
- 1 CT
Chaos Toolkitchaostoolkit.orgChaos Toolkit is a command-line tool for defining and running controlled chaos-engineering experiments against infrastructure and applications.
- 2 TToxiproxygithub.com
Free TCP proxy that alters test connections with delay, packet drops, data limits, or closed sockets.
- 3 KKrknkrkn-chaos.dev
Free Kubernetes failure injector with 20-plus scenarios, CLI, dashboard, containers, and telemetry integrations.
- 4 TTumulttumult.rs
Open-source Rust platform with 16 plugins, 91 actions, web or TOON authoring, and approval gates.
- 5 R
Reliablyreliably.comFreemium resilience platform offers 300+ actions, templates, schedules, dashboards, and experiment imports.
- 6 F
FaultMeshfaultmesh.comKubernetes-native freemium platform for controlled faults, anomaly detection, remediation, and resilience scores.
- 7 CM
Chaos Meshchaos-mesh.orgOpen-source cloud-native platform with broad fault types, Kubernetes workflows, schedules, health checks, and HTTP API.
- 8 KKubeInvaderskubeinvaders.platformengineering.it
KubeInvaders is a gamified chaos engineering and educational tool for Kubernetes that stresses a cluster so users can observe its resilience.
- 9 KAKrkn-AIgithub.com
Krkn-AI automatically runs chaos scenarios and discovers experiments to evaluate system resilience.
- 10 CChaosBladechaosblade.io
Open-source tool runs CPU, memory, disk, app, and Kubernetes experiments from CLI, HTTP, YAML, or web.
- 11 G
Gremlingremlin.comGremlin is a reliability platform that injects controlled failures and tests systems before incidents affect users.
- 12 CCCanonical Chaos Engineering Platformcanonical-chaos-engineering.readthedocs-hosted.com
Open-source Kubernetes platform deployed with Terraform and Juju, including a pod-deletion resilience probe.
- 13 PPowerfulSealpowerfulseal.github.io
Free Kubernetes failure injector with YAML scenarios for cloud, OpenStack, Kubernetes, and local systems.
- 14 PPumbagithub.com
Pumba is a chaos testing and network emulation tool for Docker, containerd, and Podman containers.
- 15 LLitmusChaoslitmuschaos.io
Litmus is an open-source chaos engineering platform for identifying infrastructure weaknesses through controlled chaos tests.
- 16 CMChaos Monkey for Spring Bootcodecentric.github.io
Chaos Monkey for Spring Boot tests Spring Boot applications by injecting failures during operation.
- 17 Cchaosdgithub.com
Free command-line or HTTP service for process, network, resource, disk, and host fault injection.
- 18 SSteadybitsteadybit.com
Paid chaos platform with a 30-day trial, no-code editor, templates, and real-time experiment view.
- 19 KMkube-monkeygithub.com
Implements Netflix’s Chaos Monkey for Kubernetes by randomly deleting pods to help develop and validate failure-resilient services.
- 20 AF
AWS Fault Injection Serviceaws.amazon.comPaid managed AWS service runs disruption experiments with tags and alarms that can stop them.
- 21 HC
Harness Chaos Engineeringharness.io - 22 KMkhaos-monkeygithub.com
Open-source Rust utility for configurable Kubernetes pod termination by count, percentage, labels, and namespaces.
- 23 AC
Azure Chaos Studioazure.microsoft.comPaid web service for scheduling experiments that inject faults into Kubernetes and cloud environments.
- 24 CMChaos Monkeygithub.com
- 25 TSTencent Smart Advisor Chaotic Fault Generatorintl.cloud.tencent.com
Places #1, #2, #3, #4, #5 on this board are open to sponsors. Take #1 · Other ways to advertise
How to choose
What does it connect to?
Start with the systems you need to disturb. Chaos Toolkit connects through an Open API and supports API, Linux, macOS, self-hosted, web, and Windows use. Chaos Mesh exposes an HTTP API and supports Pod, network, DNS, HTTP, I/O, time, stress, kernel, block-device, JVM, physical-machine, AWS, Azure, and GCP faults.
Cloud services have narrower connections. AWS Fault Injection Service runs experiments on AWS workloads and provides AWS-created scenario templates. FaultMesh uses Chaos Mesh CRDs for latency, packet loss, pod kills, CPU stress, and memory stress on a Kubernetes-native platform.
Where does it run?
Choose a deployment that matches your operations team. Chaos Toolkit, ChaosBlade, Toxiproxy, and Tumult list API, Linux, macOS, self-hosted, web, and Windows support. FaultMesh lists API, self-hosted, and web. Krkn and LitmusChaos list API, Linux, self-hosted, and web.
Managed options reduce local setup. AWS Fault Injection Service is fully managed and can run many action types without agents; instance CPU and memory utilization faults require the SSM agent. Harness Chaos Engineering lists web access with a hybrid deployment model.
What free tier or price should you expect?
Free plans cover Chaos Toolkit, FaultMesh, Krkn, Chaos Mesh, Toxiproxy, ChaosBlade, chaosd, LitmusChaos, Tumult, and Chaos Monkey for Spring Boot. Open-source options include Chaos Toolkit, ChaosBlade, LitmusChaos, Tumult, and several self-hosted utilities.
Paid terms differ. FaultMesh lists paid use from $49 per month and a 14-day trial. Reliably lists paid use from £40 per month. Gremlin and Steadybit list pricing on request with 30-day trials. AWS Fault Injection Service starts at $0.10, while Azure Chaos Studio lists no free plan.
How can you limit and stop experiments?
Use targeting and health checks to keep failures contained. Chaos Mesh supports StatusCheck resources and recurring or workflow runs. AWS Fault Injection Service uses tags to select targets and CloudWatch alarms or other tools to stop an experiment. Harness Chaos Engineering lists experiment scheduling and Kubernetes support.
Tumult adds T0–T3 risk tiers, gated approvals, quorum, TTL expiry, and content-hash pinning. Kube-monkey opts in through labels and defaults to dry run. Khaos-monkey supports fixed or percentage pod termination, namespace rules, labels, exclusions, and groups.
Which workflow suits your team?
For code-based collaboration, Chaos Toolkit stores experiments as JSON or YAML and can run in CI/CD. ChaosBlade supports CLI, Kubernetes YAML and CRDs, HTTP calls, and a web interface. Toxiproxy routes test connections through a TCP proxy and changes them over HTTP, fitting development and CI environments.
For shared operations, Reliably provides an experiment builder with more than 300 actions, templates, scheduling, scores, timelines, and dashboards. Gremlin injects faults into services, hosts, containers, and serverless workloads. Steadybit offers a no-code editor, templates, and a real-time run view.
Questions
What is a chaos engineering platform?
It injects controlled faults into software or infrastructure, then observes the response. The goal is to identify weaknesses and verify recovery behavior.
Which chaos tools have a free tier?
Chaos Toolkit, FaultMesh, Krkn, Chaos Mesh, Toxiproxy, ChaosBlade, chaosd, LitmusChaos, and Tumult list free access.
How much does FaultMesh cost?
FaultMesh lists a free plan, paid use from $49 per month, and a 14-day trial. It supports API, self-hosted, and web deployment.
Does AWS Fault Injection Service require agents?
Many AWS FIS action types run without agents. Instance-level CPU and memory utilization faults require the AWS Systems Manager agent.
Which platform offers a public API?
Chaos Toolkit has an Open API, Chaos Mesh provides an HTTP API, and chaosd offers HTTP control for its background service.
What tools support scheduled chaos experiments?
Chaos Mesh, LitmusChaos, Reliably, Harness Chaos Engineering, and Kube-monkey document scheduling or recurring experiment controls.
Some top positions on this board can be paid placements, standing where they were sold. The facts on each platform are not for sale. How we rank · Disclosure