Reliably

Overview

Reliably is a resilience-engineering platform for teams that investigate operational risks through planned experiments. It is best suited to engineering teams that want experiment design, execution and safety controls in one workflow, particularly those already using Chaos Toolkit.

Its breadth of execution options and configurable safeguards make it a flexible choice; the free plan’s one-seat limit and tight cloud-runtime and retention caps make it better for evaluation than for a growing team.

Key features

Build and manage experiments

The visual experiment builder offers more than 300 actions, including attacks and probes, without requiring users to learn a new language. Teams can import JSON Chaos Toolkit experiments and reuse custom templates, which helps those moving existing work into a managed workflow. Scheduling and review tools—including scores, freshness indicators, timelines and dashboards—give teams ways to track experiments over time.

Supported targets include AWS, Azure, Google Cloud, Kubernetes, Prometheus and Spring. Plans can run through GitHub Actions or Reliably-managed execution, or use the CLI, Docker, Kubernetes jobs or cron jobs. That range suits teams with different deployment workflows, though Reliably-managed cloud runs are subject to monthly minute allowances and per-run limits on the lower tiers.

Integrations and safety

Documented integrations include Slack, Honeycomb, Dynatrace, Google Cloud, Grafana and OpenAI ChatGPT, alongside pre-checks, safeguards and auto-pause. Slack can post experiment activity to a channel in near real time, while Honeycomb receives traces through OpenTelemetry. Environments can hold variables and encrypted secrets for use by the app, CLI or GitHub workflows.

Pre-checks can prevent an experiment from starting, and safeguards can interrupt one if configured conditions are not met. Those controls are important for teams introducing fault injection into systems where an unchecked experiment could be disruptive. Reliably describes its platform as open source and its approach as avoiding lock-in.

Pricing

Reliably uses a freemium model. The plans are billed monthly where priced; the choice comes down chiefly to team size, cloud execution needs, experiment-library limits and how long teams need to retain results.

PlanPriceLimits and inclusionsFreefreeOne team member; unlimited custom and GitHub executions; 180 cloud execution minutes per month, with a 10-minute cloud execution limit; five custom experiments; one custom template; one concurrent execution; two weeks of data retention; third-party integrations and email support.Start£40 a month, billed monthlyThree team members; unlimited custom and GitHub executions; 300 cloud execution minutes per month, with a 30-minute cloud execution limit; unlimited custom experiments; three custom templates; five concurrent executions; three months of data retention; third-party integrations and email support.Scale£330 a month, billed monthlyFive or more team members, with additional members at $20 per month; unlimited custom, GitHub and cloud executions; a 60-minute cloud execution limit; unlimited custom experiments; ten custom templates; unlimited concurrent executions; 12 months of data retention; third-party integrations and email support.EnterpriseCustom pricingUnlimited team members; unlimited custom, GitHub and cloud executions; a 60-minute cloud execution limit; unlimited custom experiments and templates; unlimited concurrent executions; 24 months of retention, unlimited with optional on-premises installation; third-party integrations and email support. An optional support SLA is available.Free is a useful single-seat starting point, but its 10-minute cloud-run cap, 180 monthly minutes and two-week retention constrain sustained work. Start raises the seat count to three and removes the experiment-count ceiling, but still caps cloud usage and concurrent runs. Scale is the more suitable paid tier for teams needing more than three seats, uncapped cloud executions and concurrency; the additional-seat charge is stated in dollars despite the plan price being in pounds. Enterprise adds unlimited seats and templates and can extend retention through on-premises installation, but carries custom pricing. Start, Scale and Enterprise specify weekday 9/5 CET support; email support is also included on Free.

Platforms

Reliably supports Linux, macOS, Windows, web and self-hosted use. Its hybrid deployment model and Kubernetes support accommodate both managed and self-managed execution, while cloud and network fault injection and blast-radius controls support resilience work across different targets.

Who it's for

Reliably is a strong fit for engineering teams that want to build and schedule resilience experiments, reuse Chaos Toolkit work, and choose between managed cloud runs and execution in their own workflows or infrastructure. It is less compelling for a solo user or small team that needs generous managed runtime and long retention at no cost: Free is limited to one member and two weeks of data, while Start costs £40 a month.

Pros and cons

Pros: More than 300 builder actions and Chaos Toolkit import support give teams multiple ways to create experiments without adopting a new experiment language.

Pros: GitHub Actions, CLI, Docker, Kubernetes jobs and cron jobs offer alternatives to Reliably-managed cloud execution.

Pros: Pre-checks and runtime safeguards can prevent or interrupt experiments when configured conditions are not met.

Cons: Free permits only one team member and limits cloud runs to 180 minutes per month and 10 minutes each, with just two weeks of retention.

Cons: Start has only three seats, three templates and five concurrent executions; teams that need unlimited concurrency or cloud executions must move to Scale.

Cons: Scale’s additional seats cost $20 per month, while Enterprise requires custom pricing.

Alternatives

Choose Chaos Toolkit instead if a free, open-source CLI is sufficient and a managed experiment workflow is unnecessary. ChaosBlade is another free, open-source option, with more than 200 experiment scenarios. Pick FaultMesh if its free Starter plan for up to three services, core fault types and basic resilience scoring is a better fit; its Growth plan costs $49 a month.

Krkn is a free alternative. Choose Toxiproxy instead for an open-source TCP proxy under the MIT license, or Chaos Mesh for an open-source, cloud-native chaos-engineering platform that runs on Kubernetes. Gremlin is a paid alternative with a 30-day free trial that does not auto-convert. Steadybit offers paid plans, including an Enterprise tier with on-premises installation, SAML, audit log and enterprise support.

For a broader comparison, see Chaos Engineering Platforms.

Verdict

Choose Reliably if your team wants a flexible way to design, schedule and execute resilience experiments with safeguards, especially if you have Chaos Toolkit work to bring along. Its main advantage is the combination of execution choices and experiment-management tools; look elsewhere if you need a generous free allowance, since the single seat, short cloud-run cap and two-week retention make Free a narrow starting point.