Buildkite vs Captain

Buildkite Pipelines is a CI/CD platform that dispatches pipeline steps as jobs to agents. Agents can be hosted by Buildkite or run in a customer's own infrastructure, including on-premises and cloud environments. Dynamic pipelines can generate and upload steps during a build, allowing a workflow to respond to code changes, test results or runtime logic. Test Engine collects results across testing frameworks and provides tools to track, analyze and act on them. It can detect and quarantine flaky tests, split tests across parallel jobs, and monitor test performance, reliability and ownership. Documented integrations include Slack, PagerDuty, Datadog, Honeycomb, OpenTelemetry, Amazon EventBridge, Artifactory, Docker Hub and Backstage. Buildkite lists TLS 1.2+ for data in transit and AES-256 encryption at rest, along with an annually audited SOC 2 Type II report. The Free Plan costs nothing and allows up to 5 users and 10 concurrent jobs. Pro costs $30 a month, billed per active user monthly, and includes up to 50 users and up to 250 concurrent agents, with 10 included.

Captain is an open-source command-line tool for finding flaky tests and managing test execution. It measures test reliability, can retry failed tests, and can quarantine flaky ones to unblock builds. By tracking test runtimes, it balances files across parallel jobs. Framework integrations include xUnit, ExUnit, Ginkgo, go test, Bun, Cucumber, Cypress, Jest, Karma, Mocha, Playwright, TestCafe, Vitest, PHPUnit, pytest, unittest, minitest, and RSpec. CI integrations listed include RWX, GitHub Actions, CircleCI, Buildkite, and GitLab. Framework results can be reviewed in a web interface for managing flakes, retries, quarantine, and suite partitioning. Feature availability varies: xUnit does not include partitioning, and Karma does not list retries or partitioning. The free OSS plan includes file-based configuration, partitioning, flaky-test tracking and quarantine, and retries. Pro costs what you agree with the maker, billed per million test results uploaded, and adds analytics on flaky tests and performance trends with 90-day retention.