Are ChatGPT or Claude better than Playwright Codegen?

The text discusses the use of AI tools like ChatGPT and Claude for generating Playwright tests. The author compares these LLMs with Playwright's built-in Codegen tool by testing them on two scenarios - searching the Playwright docs and testing a simple HTML app. They find that while vanilla ChatGPT or Claude doesn't generate good code, using an extended prompt significantly improves the resulting code. The author also notes that providing as much instruction and context as possible is crucial for obtaining high-quality results from LLMs. Furthermore, they emphasize the importance of inlining or uploading source code to receive a working test. They conclude by suggesting that AI editors like GitHub Copilot and Cursor could potentially be used to generate good Playwright code within an editor.