Which AI tools help with software QA and testing?

Start with Playwright's free test agents: a planner explores your app and writes a test plan in Markdown, a generator turns it into Playwright tests, and a healer runs the suite and repairs tests that break, driven by a coding assistant such as GitHub Copilot (free, or Pro at $10 a month) or Claude Code. TestMu AI, the new name for LambdaTest, adds KaneAI, an agent that writes end-to-end tests from plain English, from $17 a month billed annually, on a cloud of real browsers and devices. BrowserStack runs tests across real browsers and phones, with AI-driven low-code automation for testers who do not code, and plans from $12.50 a month. For a team that wants its end-to-end tests written and maintained for it, QA Wolf combines agents with a managed service. AI writes tests quickly and can write tests that pass when they should fail, so read what each one actually checks. Rule: Playwright agents if you have engineers, KaneAI or BrowserStack's low-code tools if your testers do not code.