What it does

QualityMax generates tests from a website, a repository or a written specification, then executes them in real browsers and shows the results with screenshots, video and logs. Tests can be described in plain English, and the site says one repository import produces 30 or more structured test cases with runnable Playwright or pytest scripts. When the interface changes and a selector breaks, the platform proposes a repair and shows the diff for review. It can also review pull requests and block merges through configurable quality gates in GitHub and GitLab pipelines, and trigger smoke tests from deploy webhooks. The site frames this as an independent check on AI-written code, arguing that the agent that writes code should not be the one to test it. Other pieces include qmax-code, a terminal agent that crawls, generates, runs and heals tests (the site calls it open source under Apache-2.0), a local runner called 9lives for existing Playwright suites, and MCP tools for connecting coding agents such as Claude or Codex. Plans start with a free tier that uses your own model keys.

Capabilities

Generates Playwright and pytest tests from repositories, specs and live apps
Self-healing selectors with a reviewable diff before a repair is applied
Pull request review and merge-blocking quality gates in CI
qmax-code terminal agent and MCP tools for coding agents
Web, API, performance, security and mobile testing on higher plans
Routes work across Claude, GPT and Gemini, or uses your own keys

When to use it

Your team has no dedicated QA engineer and wants tests generated and kept up to date
AI-written code is reaching your main branch and you want evidence it was checked
Your Playwright tests keep breaking when the interface changes
You want a coding agent to explore an app and draft tests through MCP