Checksum

Checksum is an AI testing platform that generates, runs and maintains automated tests for web applications and APIs. Its agents write production ready Playwright end to end tests, cover API endpoints, and run in CI on every commit and pull request. When the application changes, the agents heal broken tests automatically, and production errors are turned into new tests. The company describes the product as a background agent: a team assigns a task such as generating 200 tests and receives finished results. It is aimed at engineering teams whose code output, often from coding agents, has outpaced their ability to test it. The platform combines an E2E agent, a CI agent and an API agent that chains calls across 40 or more steps to verify state and downstream effects rather than status codes. Its Results as a Service model includes final verification by human engineers, so delivered tests are working Playwright code committed to the customer's own repository. Customers own every test and keep them if they leave. Pricing is based on the number of workflows Checksum maintains, with unlimited test runs, healing and users, and no per seat or per run fees. Three tiers are published without prices: Emerging with 50 CI ready E2E workflows and a dedicated customer engineer; Scaling with 200 workflows, custom style guides, integration with existing test infrastructure and parallel execution; and Enterprise with 400 or more workflows, the API testing agent, a custom SLA and security review support. Each tier offers a free 30 day trial, and amounts are quoted on request. Checksum positions its approach as continuous quality, an always on verification layer that runs alongside CI and CD.